Aus dem Institut für Tierzucht und Tierhaltung der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel
______
Investigations on agonistic behaviour in pigs kept under commercial farm conditions
Dissertation zur Erlangung des Doktorgrades der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel
vorgelegt von
M.Sc. Andreas Stukenborg aus Vechta
Dekanin: Prof. Dr. K. Schwarz 1. Berichterstatter: Prof. Dr. J. Krieter 2. Berichterstatter: Prof. Dr. G. Thaller
Tag der mündlichen Prüfung: 7. Februar 2011 ______Die Dissertation wurde mit dankenswerter finanzieller Unterstützung der H. Wilhelm Schaumann Stiftung angefertigt
TABLE OF CONTENTS
GENERAL INTRODUCTION ...... 1
CHAPTER ONE Agonistic behaviour after mixing in pigs under commercial farm conditions ...... 5
CHAPTER TWO Heritabilities of agonistic behavioural traits in pigs and their relationships within and between different age groups ...... 25
CHAPTER THREE Relationship between agonistic behaviour, growth and reproductive performance in pigs ...... 45
CHAPTER FOUR The use of a lesion score as an indicator for agonistic behaviour in pigs ...... 61
GENERAL DISCUSSION ...... 73
GENERAL SUMMARY ...... 81
ZUSAMMENFASSUNG ...... 83
GENERAL INTRODUCTION
Agonistic behaviour is a common component of social interactions in pigs. In general, agonistic behaviour describes both aggressive and submissive manners (Langbein and Puppe, 2004). Aggression among pigs occurs during regrouping of pigs unacquainted with one another or when competition for food is provoked by food shortage or restricted feeding space (Løvendahl et al., 2005). In common pig husbandry, agonistic behaviour can especially be observed directly after mixing when the pigs fight in order to establish a clear dominance hierarchy (Ewbank, 1976). Generally, the mixing of pigs is performed in three different age classes. The first mixing time takes place directly after weaning. The second regrouping occurs after pigs are taken from rearing to the fattening unit. The third mixing time concerns breeding sows when they are kept in a group-housing pregnancy area. Aggressive encounters often result in skin injuries and can additionally cause immunosuppressive effects (Tuchscherer and Manteuffel, 2000). Arey and Edwards (1998) reported that social stress can affect the sows’ productivity by influencing their reproductive physiology. Therefore, involvement in agonistic interactions is associated with negative effects on animal welfare as well as on the economy of pig production. An analysis of agonistic behaviour has already been made in numerous studies. However, most of these investigations have been carried out under experimental conditions and only observed pigs of one age group. The aim of the present study was to examine the agonistic behaviour of pigs over time with the aid of several behavioural traits under commercial farm conditions. Agonistic interactions were observed by video equipment at the three common mixing times in pig husbandry. The behavioural data of the three times were analysed in order to provide hints concerning the ontogenesis of agonistic behaviour.
With regard to the persistence of agonistic interactions after regrouping, Puppe et al. (1997) showed that agonistic behaviour rose continuously over several days after weaning, while other authors have observed severe fights only in the first three (Friend et al., 1983) or first 24 hours (Meese and Ewbank, 1973). Due to these inconsistent findings, the aim of Chapter One was to describe fighting behaviour over a 48-hour observation time and to expose essential observation periods in which most of the fights occur. In addition, different behavioural traits and a dominance index were presented in order to illustrate the fighting level of the pigs.
1 According to D’Eath et al. (2009), aggressive behaviour is moderately heritable and could be reduced by genetic selection. However, only a few studies have even estimated the heritabilities of agonistic behaviour after mixing pigs. For weaned piglets no heritabilities have been estimated until now. Furthermore, with relation to possible behavioural observation times for breeding it is still unexplained as to whether an aggressive female piglet becomes an aggressive growing pig and sow. Therefore, the objective of Chapter Two was to calculate correlations between the behavioural traits of different age groups. Additionally, the heritability of different agonistic behavioural traits was estimated for all three age groups.
There is also a lack of investigations in the literature which deal with the relationships between agonistic behaviour and current performance traits in pig breeding. Thus, it is still uncertain as to whether an increased consideration of agonistic behaviour in pig breeding would have negative effects on common performance traits. For this reason, Chapter Three presents correlations estimated between agonistic behavioural traits, and growth rate and backfat depth, respectively. Also relationships were analysed between reproductive performance and the agonistic behavioural traits of gilts.
A lesion score has frequently been used as a proxy indicator of post-mixing aggression (Turner et al., 2006). However, the scores mostly include an exact count of the scratches and are therefore not feasible for application under common breeding farm conditions. An easy and rapid lesion score approach is presented in Chapter Four . For this, correlations were estimated between the applied lesion score and agonistic behavioural traits in order to examine whether the lesion score could be an indicator for agonistic behaviour in pigs.
References
Arey, D.S., Edwards, S.A., 1998. Factors influencing aggression between sows after mixing and the consequences for welfare and production. Livestock Production Science 56, 61-70. D'Eath, R.B., Roehe, R., Turner, S.P., Ison, S.H., Farish, M., Jack, M.C., Lawrence, A.B., 2009. Genetics of animal temperament: aggressive behaviour at mixing is genetically associated with the response to handling in pigs. Animal 3, 1544-1554.
2 Ewbank, R., 1976. Social hierarchy in suckling and fattening pigs: A review. Livestock Production Science 3, 363-372. Friend, T.H., Knabe, D.A., Tanksley, T.D., Jr., 1983. Behavior and performance of pigs grouped by three different methods at weaning. Journal of Animal Science 57, 1406- 1411. Langbein, J., Puppe, B., 2004. Analysing dominance relationships by sociometric methods - a plea for a more standardised and precise approach in farm animals. Applied Animal Behaviour Science 87, 293-315. Løvendahl, P., Damgaard, L.H., Nielsen, B.L., Thodberg, K., Su, G., Rydhmer, L., 2005. Aggressive behaviour of sows at mixing and maternal behaviour are heritable and genetically correlated traits. Livestock Production Science 93, 73-85. Meese, G.B., Ewbank, R., 1973. The establishment and nature of the dominance hierarchy in the domesticated pig. Animal Behaviour 21, 326-334. Puppe, B., Tuchscherer, M., Tuchscherer, A., 1997. The effect of housing conditions and social environment immediately after weaning on the agonistic behaviour, neutrophil/lymphocyte ratio, and plasma glucose level in pigs. Livestock Production Science 48, 157-164. Tuchscherer, M., Manteuffel, G., 2000. The effect of psycho stress on the immune system. Another reason for pursuing animal welfare (Review). Archiv für Tierzucht - Archives of Animal Breeding 43, 547-560. Turner, S.P., Farnworth, M.J., White, I.M.S., Brotherstone, S., Mendl, M., Knap, P., Penny, P., Lawrence, A.B., 2006. The accumulation of skin lesions and their use as a predictor of individual aggressiveness in pigs. Applied Animal Behaviour Science 96, 245-259.
3
4 CHAPTER ONE
Agonistic behaviour after mixing in pigs under commercial farm conditions
Andreas Stukenborg 1, Imke Traulsen 1, Birger Puppe 2, Ulrich Presuhn 3, Joachim Krieter 1
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, D-24098 Kiel, Germany
2Leibniz Institute for Farm Animal Biology (FBN), D-18196 Dummerstorf, Germany
3farm concepts GmbH & Co. KG, D-23812 Wahlstedt, Germany
Published in Applied Animal Behaviour Science, 2011, 129: 28-35
5 Abstract
The aim of the study was to investigate agonistic behaviour of pigs after regrouping pigs under commercial sow farm conditions. The behavioural patterns were observed over a 48- hour period, directly after weaning (PIG-28; n = 647) and 40 days later (growing pigs, PIG- 68; n = 224). Agonistic interactions were analysed by noting the times (start and finish), the aggressor/receiver and the winner/loser of a fight. Differences in agonistic behaviour within and between the two age groups were recorded. The median number of fights per pig was 40.0 (PIG-28) and 6.5 (PIG-68) respectively and varied between 0 and 139 fights. A circadian rhythm for the number of fights per pig and hour was shown for both age groups. A dominance index (DI) was calculated to examine possible relations between the agonistic behaviour shown and the dominance of an individual pig. The DI was defined as the ratio of wins minus defeats divided by the sum of wins, defeats and stand-off outcomes. Dominant pigs (DI > 0) were engaged in more agonistic interactions, had a longer total fight time and initiated more fights (p < 0.05) than subordinate pigs. Significant correlations (p < 0.05) were found between the overall fight time and the dominance index with r = 0.35 (PIG-28) and r = 0.53 (PIG-68). Potential coherences between agonistic behaviour and the number of skin lesions were investigated with a lesion score (LS). The LS ranked from 0 to 4 and was determined at the beginning (LSstart) and at the end (LSend) of the observation period. The analyses of the LS showed significant preferences for the front third of the body (p < 0.05). However, unambiguous findings were only made concerning the older age group. Growing pigs with an increased LS after 48 hours had more fights per pig, a longer overall fight time and initiated more fights (p < 0.05) compared to pigs showing a lower or unchanged LS. In conclusion, the applied dominance index and lesion score are feasible methods to analyse fighting behaviour under commercial sow farm conditions. Potential relationships between agonistic behaviour and growth or reproductive performance should be analysed in further investigations. It might be beneficial to consider agonistic behaviour in pig breeding. In this context, the presented approaches are simple tools to measure agonistic pig behaviour.
Keywords: pig, agonistic behaviour, mixing, dominance index, lesion score
6 1. Introduction
Agonistic behaviour includes an aggressive as well as a submissive manner. In pig husbandry, such behaviour can especially be observed when unknown pigs are mixed into new groups. The level of aggressive behaviour depends on several circumstances such as differences in body weight, different space or group sizes or the degree of familiarity. It has been reported that heavier pigs in a group fight more with each other and win more fights (Rushen, 1987; Algers et al., 1990). Ewbank and Bryant (1972) figured out that a decrease in the area available per pig leads to an increase in agonistic interactions. Hoy and Bauer (2005) observed that mixing acquainted sows reduces the level of aggressive interactions in contrast to mixing unacquainted pigs. Aggressive encounters often result in skin injuries and can additionally cause immunosuppressive effects (Tuchscherer and Manteuffel, 2000). Hence, agonistic behaviour can lead to detrimental welfare and reduced weight gain (Tan et al., 1991; Stookey and Gonyou, 1994), thus affecting the economy of pig production. Behavioural observations had already been made in numerous studies, mostly dealing with the arising dominance relationships, often in correspondence with other behavioural and/or physiological features of the animals (Langbein and Puppe, 2004). However, these investigations had been carried out under experimental conditions, reflected by a small number of pigs in the pen. Furthermore, observations had been either made directly in the pen or with the aid of video recordings. These methods are labour-intensive and time-consuming. Recent studies (Løvendahl et al., 2005; Turner et al., 2008; Turner et al., 2009) have shown that aggressive behaviour after mixing is moderately heritable and could be reduced by genetic selection (D'Eath et al., 2009). Additionally, a lesion score has been frequently examined to find an association between aggressive behaviour and the accumulation of skin lesions. Hence, a skin lesion score assessed for the first 24 hours after regrouping is a heritable trait (Turner et al., 2006) and genetically correlated with aggressive behaviour (Turner et al., 2008; Turner et al., 2009). Therefore, skin lesions can be used to score the agonistic capability of a pig. By using a lesion score approach there is no need to make time- consuming behavioural observations. Evaluation methods, for example, could be part of the routine (Turner et al., 2008). Thus, an implementation of agonistic behaviour in commercial breeding programs would be easier to handle. In contrast to previous investigations, the data collection was carried out under commercial sow farm conditions. Consequently, the findings might prove to be more beneficial concerning the practicability of breeding pigs using agonistic behavioural traits.
7 The aim of the present paper was to show the course of agonistic behaviour in the passage of time and to examine several behavioural traits (e.g. number of fights, fight time or initiated fights per pig) at two different times of regrouping referring to common pig breeding. Thus, essential observation periods were shown, containing most of the fights and potential age discrepancies in agonistic behaviour. In addition, a dominance index and a lesion score were calculated in order to find a relationship between differences in agonistic behaviour and differences in lesion score and dominance index, respectively. Both the dominance index and the lesion score could possibly be considered simple means to measure agonistic behaviour in pigs.
2. Materials and Methods
2.1. Animals and housing
Data were collected on a closed herd sow farm of the German breeding company ‘Hülsenberger Zuchtschweine’ from October 2007 to September 2008. The herd consisted of pure-bred Large-White (LW) sows and crosses between Large-White and German Landrace (LW x DL). Due to the aim of the study to observe and analyse the behaviour of breeding sows, only female piglets from pure-bred litters (LW) and crosses from LW x DL were chosen for this study. The sows had a 28-day lactation period. Every Wednesday piglets from one sow group with 36 litters on average were weaned (one-week rhythm) and moved to the rearing house where they were sorted by gender and body size. The dimensions of the rearing pens were 5.10m x 2.10m. The pens had a fully slatted floor and were separated by 60cm high, closed partitions on the ground and two rods across. An average of 29 piglets was housed in each pen. During the rearing period the piglets were fed ad libitum with a wet feeder. Water was accessible non-stop through nipple drinkers.
2.2. Behavioural observation
Immediately after weaning four rearing pens with female breeding piglets (LW and LW x DL) were filmed for 48 hours using the HeiTel Player software program (HeiTel Digital Video GmbH, Kiel, Germany). The recording of the weaned pigs (PIG-28) started between 15:00 and 16:30 h. By exclusion of the area around the wet feeder, the entire observation area measured 3.20m x 2.10m. For the behavioural observation, ten piglets per pen were randomly chosen and individually marked on their backs. In total, the agonistic behaviour of 647 piglets
8 (PIG-28) with a total of 23,995 fights was recorded. A second observation was carried out 40 days after weaning (growing pigs = PIG-68). The same 40 pigs which had already been observed at weaning were re-marked, re-mixed and re-allocated to the four rearing pens. At the second mixing, a new group was created in equal shares from the four pens which were housed at weaning, i.e. one out of four from the new group and also two or three from the marked pigs were already known to each other. After mixing, the pigs were videotaped for 48 hours. The recording (PIG-68) started between 11:40 and 15:15 h. To reduce the effort involved, the second mixing procedure was only carried out 12 times when there were pure- bred pigs in the rearing pens. These pigs were of particular importance for further studies because of their existing pedigrees. However, due to problems with the technical supply, only 224 pigs (PIG-68) (2,024 fights) were completely observed for the entire 48 hours. The agonistic behaviour of the marked pigs was analysed with the aid of the videotapes. The growing pigs were evaluated by one person. Four observers recorded the agonistic interactions for the weaned piglets. Prior to visual assessment, the observers were trained in the definition and identification of agonistic behaviour. Characteristically distinctive fighting sequences were studied to standardise the ratings of the fights. Finally, they were tested with the same training videotape. The derived data were stored in a database. Thereby, an agonistic interaction was defined as a fight or a displacement with physical contact initiated by one individual and featuring aggressive behavioural elements followed by any form of submissive behaviour performed by the opponent (Langbein and Puppe, 2004). Agonistic behaviour was recorded when the fights took longer than one second. If there was an intervening period of more than eight seconds, a new fight was considered to have been started (Puppe, 1998). In accordance with Puppe (1998), fights were displayed in the form of bodily attacks such as ‘head-to-head knocks’ and ‘head-to-body knocks’, ‘parallel/inverse parallel pressings’, ‘bitings’ or ‘physical displacements’ which were induced by the aggressor pig. The loser of a fight was defined as the animal showing a submissive manner, e.g. it stopped fighting, turned away from an attack, tried to flee or was displaced from the location (Tuchscherer et al., 1998; Langbein and Puppe, 2004). The times (start time and finish time), the aggressor/receiver and the winner/loser of a fight were noted for each fight in which a marked pig was involved. If there was no clear outcome, the fight was designated as a stand-off. If an agonistic interaction moved out of the observation area for more than one minute, the fight was defined as terminated and the outcome was “unclear”.
9 2.3. Dominance index
A dominance index (DI) was calculated to express the strength and the dominance of the observed pigs. The index was chosen referring to Bowen and Brooks (1978), who designed an index as the ratio of wins minus defeats to all decisive fights. However, due to the high number of stand-offs for a considerable number of piglets in the present study, the DI was calculated including the fights which were classed as indecisive. In this way, the dominance of a “stand-off” piglet was reflected more accurately. The DI was defined as the sum of wins minus defeats to wins, defeats and stand-off outcomes. = wins −defeats DI ∑ wins +defeats +stan doffs According to this calculation, no DI can be calculated for those pigs that were not involved in any fighting or only had uncertain fight outcomes (one for PIG-28, 15 for PIG-68). The dominance index ranked from -1 (absolute submissive) to +1 (absolute dominant). In order to plot the DI (Figure 2), values were summarised in steps of 0.1 (0 = -0.05 to 0.05), except the DI from 0.9 and -0.9, respectively, which included all data from 0.85 (-0.85) to <1 (> -1).
2.4. Lesion score
In addition to the behavioural observations, a skin lesion score (LS) was assigned to all pigs marked immediately before starting the 48-hour observation period (LSstart). After 48 hours (PIG-68) or one week later (PIG-28), the LS was recorded again (LSend) for the majority of the observed pigs. The time differences between the first and second recordings (48 hours vs. one week) were due to management reasons (the workflow on the farm was to be disrupted as little as possible). To evaluate the skin areas mainly affected, the body was divided into three regions: front (head, neck, shoulders and front legs), middle (flanks and back) and rear (rump, hind legs and tail) (modified according to Turner et al., (2006)). The three regions were evaluated independently of each other. The LS ranked from 0 (no wounds and scratches) to 4 (many, deep wounds and scratches). The change in the LS (LSdiff) was described as the difference of LSend minus LSstart. Based on this definition the LSdiff could also be negative for pigs showing more scratches at the beginning than at the end of the observation period. In total, the LSdiff was calculated for 606 (PIG-28) and 184 (PIG-68) pigs, respectively.
10 2.5. Statistical analysis
Data analysis was performed using the SAS statistical software package (SAS Institute Inc., 2004). The UNIVARIATE procedure in SAS (2004) was applied to test for normal distribution (Shapiro-Wilk test). Due to the non-normal distribution of all behavioural traits, including the fights per pig and hour, the medians were calculated for descriptive statistics (Proc MEANS; SAS 2004). The Wilcoxon rank-sum test (Proc NPAR1WAY; SAS 2004) was used to identify significant differences in the number of agonistic interactions per hour during daytime. Medians were compared between morning, midday and evening as well as the night and daytime results. The same test was applied to compare the medians of the behavioural parameters from animals with a DI larger or less than zero. Pigs with a DI = 0 were not considered in this test. Previous investigations had shown that these pigs did not influence the results. Furthermore, the Wilcoxon rank-sum test was used to find differences between the LS of the three body areas and to compare the medians of behavioural parameters from pigs with a positive or non-positive front LSdiff. Due to the non-normal distribution of the different behavioural traits, Spearman correlation coefficients (Proc CORR; SAS, 2004) were calculated for behavioural parameters, dominance index and the front LSdiff.
3. Results
3.1. Behavioural parameters
Weaned piglets (PIG-28) A winner and a looser could be assigned to 39.1% of all the observed agonistic interactions. Most fights were rated as stand-offs (47.8%). Due to the existing space outside the observation area, the remaining outcomes (13.1%) and additionally 23.7% of the fight beginnings could not be specified. Table 1 shows the different behavioural indicators. All traits presented a wide range. One piglet had no fight at all in the first 48 hours; in contrast, one piglet had 139 agonistic interactions. The number of fights per piglet and hour is illustrated in Figure 1. The course shows significant variations during the times of day. In general the number of agonistic interactions was significantly smaller during the night (h 2-13 and h 26-37) than during the daytime (h 1, h 14-25, and h 38-48) (p < 0.05).
11 Table 1: Median, minimum (Min.) and maximum (Max.) of the behavioural parameters, observed for 48 hours after weaning (PIG-28) and second mixing (PIG68) PIG-28 (n = 23,995) PIG-68 (n = 2,024) Median Min. Max. Median Min. Max. Fights per pen 1) 443 191 783 96 28 152 Fights per pig 40 0 139 6.5 0 86 Fight time per pig (s) 2471 0 13171 87 0 6751 Fighting duration (s) 29 1 1114 12 1 1299 Initiated fights per pig 14 0 90 2 0 44 Fights won per pig 6 0 55 0 0 41 Fights lost per pig 7 0 63 2 0 15 Stand-off fights per pig 20 0 70 3 0 33 1) observation area
2.5 PIG-28 (n = 23995) midday b midday b morning a 2 evening a morning a
1.5 evening a 1
0.5 fights/piglet/hour fights/piglet/hour
0 01 06 12 18 24 30 36 42 48
1.5 PIG-68 (n = 2024)
a c b 1 evening b midday morning morning evening a 0.5
0 fights/pig/hour 01 06 12 18 24 30 36 42 48 fights/pig/hour hours after mixing
Figure 1: Fights per piglet and hour after the first (PIG-28) and second (PIG-68) mixings. Daytimes with different letters are significantly different (p ≤ 0.05)
12 Growing pigs (PIG-68) For the growing pigs, 47.1% of the agonistic interactions showed a clear result. 43.7% of the fights were observed as stand-offs and 9.2% had no outcome in the observation area. The initiator was out of sight in 17.2% of the fights. Similar to the younger pigs, the behavioural traits of these growing pigs showed a wide range of results, but the mean values were considerably lower (Table 1). In total, 13 pigs had no fights in the first 48 hours. The growing pigs fought much more during the daytime than during the night (p < 0.05) (Figure 1). No agonistic interaction was observed in hour 38.
3.2. Dominance index
Weaned piglets (PIG-28) More than two thirds of the piglets had a DI of between -0.25 and +0.25. There was one absolutely submissive piglet (DI = -1) and nine animals had a high index of 0.6 (see Figure 2). The median DI was -0.02.
120
100 PIG-28 (n = 647) PIG-68 (n = 224) 80
60
40 Number Number of pigs
20
0 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 Dominance index
Figure 2: Dominance index of weaned piglets (PIG-28) and growing pigs (PIG-68)
Growing pigs (PIG-68) The median DI for the growing pigs was -0.33. Twenty-six animals had no winner outcomes (DI = -1) in contrast to one pig which was absolutely dominant. The other dominant pigs had an index lower than 0.65. Fifteen of the 224 pigs had no DI. These pigs had no fights or the fight outcomes were uncertain.
13 Comparison of dominant vs. submissive piglets/pigs A comparison of dominant (DI > 0) and subordinate (DI < 0) pigs showed significant differences (p < 0.05) for the major behavioural traits in both age groups (Table 2). The dominant piglets were more active than the subordinates, expressed by a higher number of fights and a longer time involved in agonistic interactions (p < 0.05). In addition, piglets with a positive DI initiated more fights (p < 0.05). Furthermore, they had considerably more wins than defeats (p < 0.05), although this was due to the definition of the index.
Table 2: Comparison of median behavioural traits between dominant and submissive pigs for the two age groups (PIG-28/PIG-68) PIG-28 (n = 23,995) PIG-68 (n = 2,024) DI < 0 DI > 0 DI < 0 DI > 0 Fights per pig 37 a* 45 b 5a 18 b Fight time per pig (s) 1817 a 3290 b 56 a 558 b Initiated fights per pig 12 a 17 b 1a 7b Fights won per pig 3a 10 b 0a 6b Fights lost per pig 10 a 3b 2a 1.5 a *Within rows and within age groups (PIG-28/PIG-68) medians with different letters are significantly different (Wilcoxon rank-um test; p < 0.05)
3.3. Lesion score
Weaned piglets (PIG-28) Figure 3 shows the differences in the LS for the three body areas. No positive LSdiff, equivalent to no additional wounds in the observation period, was found for almost 51% (middle) and 68% (rear) of the piglets, respectively. More than 70% of the animals had a positive LSdiff for the front body third. One piglet exhibited a difference of four units. At the second LS recording, the LS for the front body area was significantly higher than the LS for the middle and rear body area (p < 0.05). Negative LSdiff was especially shown for the rear body region (ca. 14%). Thus, these piglets had more wounds at the first than at the second LS recording. In contrast, only about 5% of the front and 7% of the middle body third had a negative LSdiff.
14 350 PIG-28 (n = 647) 300
250
200 front middle 150 rear
Number of piglets 100
50
0 -2 -1 0 1 2 3 4 LS difference 120 PIG-68 (n = 224)
100
80 front 60 middle rear 40 Number Number of pigs
20
0 -2 -1 0 1 2 3 4
LS difference Figure 3: Lesion score difference of weaned piglets (PIG-28) and growing pigs (PIG-68) for the front, middle and rear body areas
Growing pigs (PIG-68) Except for the anterior third, the distribution of the LSdiff was similar to that of the younger piglets (Figure 3). A considerable number of pigs showed only a small increase in the LS. No positive LSdiff was shown for approximately 62% and 72%, respectively, with regard to the middle and rear regions. About 56% had an increased LS after mixing, regarding the front third. Moreover, the rear and middle body area had a significantly smaller LS than the front region (p < 0.05).Due to these proportions, the front third was clearly preferred.
15 Comparison of negative/unchanged vs. positive LSdiff The behavioural traits of piglets with new wounds were compared to animals without new scratches. Due to the predominant body area, only the front region was considered. The young pigs had an equal number of fights per piglet (Table 3). Significant differences (p < 0.05) were shown in the fight time and the number of defeats. For the older pigs only the number of defeats was not significant.
Table 3: Comparison of median front lesion score difference (LSdiff) (negative/unchanged vs. positive LSdiff) for the two age groups (PIG-28/PIG-68) PIG-28 (n = 23,995) PIG-68 (n = 2,024) LSdiff ≤ 0 LSdiff > 0 LSdiff ≤ 0 LSdiff > 0 Fights per pig 39 a* 41 a 3a 9b Fight time per pig (s) 1915 a 2757 b 35.5 a 174 b Initiated fights per pig 14 a 14 a 1a 2b Fights won per pig 6a 6a 0a 0b Fights lost per pig 8a 6b 2a 2a Average DI -0.05 a -0.01 a -0.5a -0.12 b *Within rows and within age groups (PIG-28/PIG-68) median with different letters are significantly different (Wilcoxon rank-sum test; p < 0.05)
3.4 Correlation between dominance index, the front LSdiff and behavioural parameters
Table 4 shows the correlations (Spearman correlation coefficients) between behavioural parameters, dominance index and the lesion score difference. The dominance index was correlated with all observed behavioural traits (p < 0.05). The older pig group had apparently higher correlation coefficients, especially for the fights per piglet, the fight time per piglet and the initiated fights. In contrast to the number of fights won, the lost fights for the younger piglets were negatively correlated with the DI. However, their correlation with the DI was on the same level (r = 0.64/-0.60). For the older pigs, the fights won were highly correlated with the DI (r = 0.74), whereby the DI and the fights lost showed an apparently smaller correlation (r = -0.31). With regard to the younger animals, the LSdiff was not correlated with the number of wins and the number of initiated fights. A slightly negative correlation was found between the LSdiff and the number of fights lost. No correlation was detected between the LSdiff and the number of fights lost with regard to the older pigs. Otherwise, the LSdiff was highly correlated with all other parameters.
16 Table 4: Correlation between the front lesion score difference (LSdiff), dominance index (DI) and behavioural parameters PIG-28 (n = 23,995) PIG-68 (n = 2,024) DI LSdiff DI LSdiff Fights per pig 0.18* 0.10* 0.47* 0.36* Fight time per pig 0.35* 0.32* 0.53* 0.42* Initiated fights per pig 0.19* 0.02 0.57* 0.29* Fights won per pig 0.64* 0.06 0.74* 0.39* Fights lost per pig -0.60* -0.13* -0.31* 0.07
DI - 0.13* - 0.38* *p < 0.05
4. Discussion
Previous investigations have already shown that agonistic behaviour in pigs has a wide phenotypic range (Jensen, 1982; Forkman et al., 1995; Otten et al., 1997; Puppe et al., 1997). The current data support this statement. Moreover, the results of the present paper illustrate a clear circadian rhythm in agonistic behaviour with high aggression frequencies during daytime. Age differences between weaned and growing pigs in the number of performed agonistic interactions were ascertained. Furthermore, the findings indicate that both the applied DI and the LS approach (except for the weaned piglets) could be used to make basic statements concerning the aggressiveness of pigs.
Behavioural parameters A low increase in fighting activity was observed for the weaned piglets directly after mixing. The growing pigs had the highest number of fights in the first hour after regrouping. Because of the novel situation with a great number of new pen mates, it might be that the first few hours after mixing are always fight times of high frequency regardless of the time of day. However, taking the whole 48-hours observation period into account a circadian rhythm is clearly visible. This is in accordance with Hessel et al. (2006), who observed a declined rate in agonistic behaviour at night. McGlone and Newby (1994) proved a circadian pattern in activity independent of group size. In respect of further investigations, the presence of a circadian rhythm could decrease the amount of time and work effort incurred by using videotapes. Potentially, recording is only required during the daytime when the majority of fights occur. Therefore it still has to be clarified as to whether there are pigs which are also (or
17 only) nocturnal or if all pigs fall into this circadian pattern. Puppe et al. (1997) showed that agonistic behaviour in weaned piglets rose continuously to day 4 after weaning. While Friend et al. (1983) observed severe fights only in the first three hours after mixing but not after 24 and 48 hours. Meese and Ewbank (1973) suggested that vigorous fighting was virtually eliminated after 24 hours and the social order was fixed within 48 hours. The current time courses in the study presented here showed no considerable decrease in the fighting level over the whole 48-hour observation period. Hence, a 24-hour observation would ignore a great quantity of information. On the contrary, for an exact description of agonistic behaviour, it might be useful to observe the pigs for considerably more than 48 hours. An observation for three or even four days after weaning could be appropriate. The present study shows a considerable difference in the quantity of fights between the groups of younger (PIG-28) and older pigs (PIG-68). A possible reason for the differences in the level of aggression could be the higher degree of familiarity inside a pen of growing pigs compared to weaned piglets. Due to the mixing procedure it became obvious that one out of four pigs already knew each other after the second regrouping. In this regard, Puppe’s (1998) observation covered 12-week-old pigs in the pen area and concluded that familiar pigs were engaged in fewer agonistic interactions than unfamiliar pigs. Jensen and Yngvesson (1998) found no significant difference in the fighting behaviour between unknown and pre-exposure pigs. In their opinion, it is difficult to inhibit overt fighting between newly mixed pigs through pre-exposure. However, concerning the high agonistic level of the younger piglets in the present study, various authors have indicated that socialising pigs before weaning subsequently leads to a restricted level of fighting (Weary et al., 1999; Pitts et al., 2000; Hessel et al., 2006; Parratt et al., 2006). Such practice would reduce the number of stressors and improve the welfare of weaned piglets, which are also influenced by a new environment and diet and the separation from the sow (D'Eath, 2005). Nevertheless, pre-weaning mixing times or even weaning without separating the litters has yet been not implemented in common pig husbandry. Currently, pigs are sorted by sex as in the present study, which also seems to have a decreasing effect on agonistic behaviour (Colson et al., 2006). There were also differences between the two age groups with regard to their space allowance. Pigs (PIG-68) with a lower space allowance (kg/m 2) had a clearly lower aggression level overall compared with piglets with more space allowance (PIG-28). Based on a live weight of 28 kg for the older pigs (PIG-68), the space allowance was approximately 76 kg/m 2 with about 2.7 pigs per m2. This stocking density may lead to restricted movement and social behaviour, which in turn induces fewer agonistic interactions. Meunier-Salaun et al. (1987) described that
18 crowding induces a breakdown in communications and pigs adopt a “social avoidance” strategy, preventing aggression escalation. Whereas, according to Ewbank and Bryant (1972), decreasing the area available per pig seems to increase the number of agonistic interactions.
Dominance index The distribution and the median value of about zero with regard to the DI from the young piglets (PIG-28) indicates that there was a uniform ratio between subordinates and dominants in the pen. In contrast, after the second mixing, a median value of -0.33 indicated that most of the growing pigs had a negative DI. These various allocations could be an indication of a changing dominance hierarchy among the animals over time. Presumably, there were only a few highly dominant growing pigs in the pen. Since nearly two thirds of the pigs were unmarked, the majority of the few dominant animals belonged to this group. A marking of all pigs in the pen would have illustrated a more precise dominance hierarchy. Based on the fact that the analysis was carried out in normal rearing pens, a marking of all 30 pigs in the pen was not possible due to the considerable time requirement and the decreasing overview. However, Table 4 shows a considerable number of correlations between the DI and the observed behavioural traits. Furthermore, clearly significant differences in the behavioural parameters were shown between the dominant and subordinate pigs, although using the least possible distinction (DI > 0 or DI < 0). These significant differences showed that the animals can be divided into more or less active fractions with the aid of a dominance index, regarding their agonistic behaviour.
Lesion score According to McGlone (1985), the lesion score clearly shows high stress for the front body third in contrast to the middle and especially to the posterior third. In the present study, the younger piglets (PIG-28) had a higher front LSdiff than older pigs (PIG-68). This reflected the higher involvement in agonistic interactions of the young piglets. Associated with these findings, an investigation of the rear and of the middle lesion score seemed to be rather insignificant in relation to agonistic behaviour. Significant correlations were found between the LSdiff and the behavioural parameters for both age groups. However concerning the weaned piglets, the correlations were only small and only significant for three out of five behavioural traits. Thus, a general connection between these parameters and the LS could not be demonstrated. For the growing pigs, all behavioural parameters, except the number of fights lost, were significantly correlated with
19 the LSdiff. In addition, the correlation coefficients were higher compared to the younger age group (Table 4). Relationships between the LSdiff and the behavioural parameters were also shown by the comparison of the positive and non-positive LSdiff groups (Table 3). These findings confirmed the results of Turner et al. (2006), who calculated a positive correlation between the front LS and the proportion of time spent in reciprocal fighting for growing pigs. The weaned piglets fought much more than the growing pigs but apparently their interactions were not as intensive, so a fight did not always end in injury. In this respect, Donaldson et al. (2002) suggested that social play behaviour could also occur during serious aggression. Thus, the LS approach presented allowed no clear indication of agonistic interactions in weaned piglets. In contrast, the growing pigs showed a more obvious aggressiveness. Therefore, the agonistic interaction mostly ends in injury, which means that a greater number of fights also results in a higher LS. According to this, the LS method can be used as a means to estimate agonistic behaviour in growing pigs. It should be mentioned that there are references indicating that a body weight asymmetry within a pen is positively correlated with the number of scratches (Olesen et al., 1996; Turner et al., 2006; Li and Johnston, 2007). However, the statements concerning the relation between behavioural traits and pig live-weight are not consistent. D’Eath (2002) reported that heavier pigs had an overall longer fight time, initiated more fights and won more often than lighter pigs. In contrast, Turner et al. (2006) found no correlation between body weight and fight time. The overall level of agonistic interactions seemed to be lower in pens with obvious differences in body weight, whereas heavier pigs fought more among themselves resulting in a high number of body lesions (Rushen, 1987; Andersen et al., 2000). However, because of the difficulties of implementing the weighing procedure in the operation flow of a sow farm, live weight was not recorded. This fact makes clear that data recording under common conditions is often associated with undesirable restrictions.
5. Conclusion
In order to assess an exact description of the agonistic potential of pigs, it might be useful to observe agonistic behaviour for more than 48 hours after mixing. It may be sufficient to observe the animals only in the daytime, when the majority of fights occur. Furthermore, pigs can be divided into more or less aggressive fractions with the aid of a dominance index and the lesion score approach is a relatively easy and simple means to estimate agonistic behaviour in growing pigs. In general, the present paper reveals that it is possible to analyse
20 agonistic behaviour under commercial farm conditions. As such, the applied methods could be incorporated into further investigations to measure and estimate agonistic interactions. This in turn would finally help to improve the animals’ welfare.
References
Algers, B., Jensen, P., Steinwall, L., 1990. Behaviour and weight changes at weaning and regrouping of pigs in relation to teat quality. Applied Animal Behaviour Science 26, 143-155. Andersen, I.L., Andenæs, H., Bøe, K.E., Jensen, P., Bakken, M., 2000. The effects of weight asymmetry and resource distribution on aggression in groups of unacquainted pigs. Applied Animal Behaviour Science 68, 107-120. Bowen, D.W., Brooks, R.J., 1978. Social organization of confined male collared lemmings (Dicrostonyx groenlandicus Traill). Animal Behaviour 26, 1126-1135. Colson, V., Orgeur, P., Courboulay, V., Dantec, S., Foury, A., Mormède, P., 2006. Grouping piglets by sex at weaning reduces aggressive behaviour. Applied Animal Behaviour Science 97, 152-171. D'Eath, R.B., 2002. Individual aggressiveness measured in a resident-intruder test predicts the persistence of aggressive behaviour and weight gain of young pigs after mixing. Applied Animal Behaviour Science 77, 267-283. D'Eath, R.B., 2005. Socialising piglets before weaning improves social hierarchy formation when pigs are mixed post-weaning. Applied Animal Behaviour Science 93, 199-211. D'Eath, R.B., Roehe, R., Turner, S.P., Ison, S.H., Farish, M., Jack, M.C., Lawrence, A.B., 2009. Genetics of animal temperament: aggressive behaviour at mixing is genetically associated with the response to handling in pigs. Animal 3, 1544-1554. Donaldson, T.M., Newberry, R.C., Špinka, M., Cloutier, S., 2002. Effects of early play experience on play behaviour of piglets after weaning. Applied Animal Behaviour Science 79, 221-231. Ewbank, R., Bryant, M.J., 1972. Aggressive behaviour amongst groups of domesticated pigs kept at various stocking rates. Animal Behaviour 20, 21-28. Forkman, B., Furuhaug, I.L., Jensen, P., 1995. Personality, coping patterns and aggression in piglets. Applied Animal Behaviour Science 45, 31-42.
21 Friend, T.H., Knabe, D.A., Tanksley, T.D., Jr., 1983. Behavior and performance of pigs grouped by three different methods at weaning. Journal of Animal Science 57, 1406- 1411. Hessel, E.F., Reiners, K., Van den Weghe, H.F.A., 2006. Socializing piglets before weaning: Effects on behavior of lactating sows, pre- and postweaning behavior, and performance of piglets. Journal of Animal Science 84, 2847-2855. Hoy, S., Bauer, J., 2005. Dominance relationships between sows dependent on the time interval between separation and reunion. Applied Animal Behaviour Science 90, 21- 30. Jensen, P., 1982. An analysis of agonistic interaction patterns in group-housed dry sows - Aggression regulation through an avoidance order. Appl. Anim. Ethol. 9, 47-61. Jensen, P., Yngvesson, J., 1998. Aggression between unacquainted pigs - sequential assessment and effects of familiarity and weight. Applied Animal Behaviour Science 58, 49-61. Langbein, J., Puppe, B., 2004. Analysing dominance relationships by sociometric methods - a plea for a more standardised and precise approach in farm animals. Applied Animal Behaviour Science 87, 293-315. Li, Y.Z., Johnston, L.J., 2007. Effects of familiarity and weight variation on aggression among grow-finish pigs following regrouping. Journal of Animal Science 85 (Suppl. 2), 49. Løvendahl, P., Damgaard, L.H., Nielsen, B.L., Thodberg, K., Su, G., Rydhmer, L., 2005. Aggressive behaviour of sows at mixing and maternal behaviour are heritable and genetically correlated traits. Livestock Production Science 93, 73-85. McGlone, J.J., 1985. A quantitative ethogram of aggressive and submissive behaviors in recently regrouped pigs. Journal of Animal Science 61, 559-565. McGlone, J.J., Newby, B.E., 1994. Space requirements for finishing pigs in confinement: behavior and performance while group size and space vary. Applied Animal Behaviour Science 39, 331-338. Meese, G.B., Ewbank, R., 1973. The establishment and nature of the dominance hierarchy in the domesticated pig. Animal Behaviour 21, 326-334. Meunier-Salaun, M.C., Vantrimponte, M.N., Raab, A., Dantzer, R., 1987. Effect of floor area restriction upon performance, behavior and physiology of growing-finishing pigs. Journal of Animal Science 64, 1371-1377.
22 Olesen, L.S., Nygaard, C.M., Friend, T.H., Bushong, D., Knabe, D.A., Vestergaard, K.S., Vaughan, R.K., 1996. Effect of partitioning pens on aggressive behavior of pigs regrouped at weaning. Applied Animal Behaviour Science 46, 167-174. Otten, W., Puppe, B., Stabenow, B., Kanitz, E., Schön, P.C., Brüssow, K.P., Nürnberg, G., 1997. Agonistic interactions and physiological reactions of top- and bottom-ranking pigs confronted with a familiar and an unfamiliar group: Preliminary results. Applied Animal Behaviour Science 55, 79-90. Parratt, C.A., Chapman, K.J., Turner, C., Jones, P.H., Mendl, M.T., Miller, B.G., 2006. The fighting behaviour of piglets mixed before and after weaning in the presence or absence of a sow. Applied Animal Behaviour Science 101, 54-67. Pitts, A.D., Weary, D.M., Pajor, E.A., Fraser, D., 2000. Mixing at young ages reduces fighting in unacquainted domestic pigs. Applied Animal Behaviour Science 68, 191- 197. Puppe, B., 1998. Effects of familiarity and relatedness on agonistic pair relationships in newly mixed domestic pigs. Applied Animal Behaviour Science 58, 233-239. Puppe, B., Tuchscherer, M., Tuchscherer, A., 1997. The effect of housing conditions and social environment immediately after weaning on the agonistic behaviour, neutrophil/lymphocyte ratio, and plasma glucose level in pigs. Livestock Production Science 48, 157-164. Rushen, J., 1987. A difference in weight reduces fighting when unacquainted newly weaned pigs first meet. Canadian Journal of Animal Science 67, 951-960. SAS Institute Inc., 2004. SAS/STAT User's guide, Version 9.1. Stookey, J.M., Gonyou, H.W., 1994. The effects of regrouping on behavioral and production parameters in finishing swine. Journal of Animal Science 72, 2804-2811. Tan, S.S.L., Shackleton, D.M., Beames, R.M., 1991. The effect of mixing unfamiliar individuals on the growth and production of finishing pigs. Animal Production 52, 201-206. Tuchscherer, M., Manteuffel, G., 2000. The effect of psycho stress on the immune system. Another reason for pursuing animal welfare (Review). Archiv für Tierzucht - Archives of Animal Breeding 43, 547-560. Tuchscherer, M., Puppe, B., Tuchscherer, A., Kanitz, E., 1998. Effects of social status after mixing on immune, metabolic, and endocrine responses in pigs. Physiology & Behavior 64, 353-360.
23 Turner, S.P., Farnworth, M.J., White, I.M.S., Brotherstone, S., Mendl, M., Knap, P., Penny, P., Lawrence, A.B., 2006. The accumulation of skin lesions and their use as a predictor of individual aggressiveness in pigs. Applied Animal Behaviour Science 96, 245-259. Turner, S.P., Roehe, R., D'Eath, R.B., Ison, S.H., Farish, M., Jack, M.C., Lundeheim, N., Rydhmer, L., Lawrence, A.B., 2009. Genetic validation of postmixing skin injuries in pigs as an indicator of aggressiveness and the relationship with injuries under more stable social conditions. Journal of Animal Science 87, 3076-3082. Turner, S.P., Roehe, R., Mekkawy, W., Farnworth, M.J., Knap, P.W., Lawrence, A.B., 2008. Bayesian analysis of genetic associations of skin lesions and behavioural traits to identify genetic components of individual aggressiveness in pigs. Behavior Genetics 38, 67-75. Weary, D.M., Pajor, E.A., Bonenfant, M., Ross, S.K., Fraser, D., Kramer, D.L., 1999. Alternative housing for sows and litters: 2. Effects of a communal piglet area on pre- and post-weaning behaviour and performance. Applied Animal Behaviour Science 65, 123-135.
24 CHAPTER TWO
Heritabilities of agonistic behavioural traits in pigs and their relationships within and between different age groups
Andreas Stukenborg 1, Imke Traulsen 1; Eckhard Stamer 2, Birger Puppe 3, Ulrich Presuhn 4, Joachim Krieter 1
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, D-24098 Kiel, Germany
2TiDa Tier und Daten GmbH, D-24259 Westensee/Brux, Germany
3Leibniz Institute for Farm Animal Biology (FBN), D-18196 Dummerstorf, Germany
4farm concepts GmbH & Co. KG, D-23812 Wahlstedt, Germany
25 Abstract
The aim of the present study was to estimate heritabilities for different agonistic behavioural traits after mixing pigs of three age classes and to compare the agonistic behaviour between the age groups. The behavioural patterns were observed over a 48-hour period, directly after weaning (PIG-28; n = 884), 40 days later (growing pigs, PIG-68; n = 351) and immediately after mixing gilts (GILTS; n = 389). Agonistic interactions were analysed by noting the times (start and finish), the aggressor/receiver and the winner/loser of a fight. Starting from this, the number and the time involved were calculated for ten agonistic behavioural traits. Intermediate heritabilities were estimated for growing pigs, with the highest heritability for the number of won fights (h 2 = 0.37). In comparison, the heritabilities for weaned piglets were at a lower level with the highest heritability for the duration spent in initiated fights (h 2 = 0.20). Concerning gilts, substantial heritabilities were only found for the number of received aggressions (h 2 = 0.42), the time spent in received aggressions (h2 = 0.15), the number of lost fights (h 2 = 0.13) and the time spent in lost fights (h 2 = 0.09). Within an age group, the number of initiated fights was always more highly correlated with the number of won fights than with the number of lost fights indicating that pigs which initiated a fight more often also won fights more often. Moderate correlations were found between the behavioural traits from growing pigs and gilts, while relationships between the weaned piglets and older age groups were only small. In conclusion, the results indicate that more aggressive growing pigs tended to also be more aggressive sows. Additionally, the agonistic behaviour from weaned piglets cannot be compared with other age groups since the agonistic behaviour of weaned piglets is probably related to playfulness while older pigs seem to establish a new dominance hierarchy.
Keywords: pig, agonistic behaviour, heritability, correlation, age groups
26 1. Introduction
In common pig husbandry, agonistic behaviour especially occurs when unacquainted pigs are mixed with others groups of acquainted pigs in order to establish a clear dominance hierarchy (Ewbank, 1976). Mixing is usually performed in three different age classes: the first directly after weaning; the second time after the pigs are taken from rearing into growing pens and finally, a third regrouping concerns breeding sows, when they are kept in a group-housing pregnancy area. Agonistic behaviour is often accompanied by negative effects on pig production. For example, pigs with a high level of aggression require more energy (van Erp-van der Kooij et al., 2003). The aggressive encounters often result in skin injuries and can additionally cause immunosuppressive effects (Tuchscherer and Manteuffel, 2000). As a result, agonistic behaviour can be detrimental to pig welfare and also affects the economy of pig production by reduced weight gain (Tan et al., 1991; Stookey and Gonyou, 1994). Until now, aggressive behaviour has hardly been considered in pig breeding programs. Only a few studies have previously estimated the heritabilities of agonistic behaviour after mixing pigs. Løvendahl et al. (2005) observed aggressive behaviour of sows at mixing and calculated intermediate heritabilities for performed aggression (h 2 = 0.17 to 0.24) but lower estimates for received aggression (h 2 = 0.04 to 0.06). Heritabilities were found for growing pigs, ranging from h 2 = 0.08 to 0.43 (Turner et al., 2009) and h 2 = 0.17 to 0.46 (Turner et al., 2008), respectively. As yet no heritability estimates has been calculated for weaned piglets. Nevertheless, it can be assumed that aggressive behaviour is moderately heritable and could be reduced by genetic selection (D'Eath et al., 2009). With regard to possible selection times for breeding, it is still arguable as to whether aggressive female piglets become aggressive growing pigs and sows. The correlations between the three common mixing times would probably provide information about the ontogenesis of agonistic behaviour. The aim of the present study was to analyse agonistic behaviour in three age classes under commercial farm conditions. Heritability estimates for the defined agonistic behavioural traits were shown. Finally, correlations were calculated between agonistic behavioural traits within an age group as well as within a trait between the age groups.
27 2. Material and Methods
2.1. Animals and Housing
Data were collected on a closed herd sow farm of the German breeding company ‘Hülsenberger Zuchtschweine’ from October 2007 to April 2009. The herd composed of pure- bred Large-White (LW) sows and crosses between Large-White and German Landrace (LW x DL). Female pigs from pure-bred litters (LW) and crosses from LW x DL were chosen for this study. The sows were managed in a one-week rhythm and had a 28-day lactating period. Every Wednesday, piglets from 36 litters were weaned and moved to the rearing house where they were sorted by gender and body size. For the presented study, the pig behaviour was observed (see section 2.2) in three age groups: weaned piglets (PIG-28), growing pigs (PIG- 68) and gilts (GILTS). The housing conditions during the three observation times are described below.
PIG-28/PIG-68 After weaning, 29 piglets on average were housed in a rearing pen. The dimensions of the rearing pens were 5.10m x 2.10m. The pens had a fully slatted floor and were separated by 60cm high, closed partitions on the ground and two rods across. During the rearing period the piglets were fed ad libitum with a wet feeder. Water was accessible non-stop via nipple drinkers. Both weaned piglets and growing pigs were observed in the rearing pens.
GILTS The gilts were kept as an individual group in the pregnancy area. Before the gilts were placed in their pen, all animals were washed and were given new earmarks for the feeding station. On average, 48 gilts were housed in the gilt group. The dimension of the gilt pen was ca. 15.5m x 6.5m. The pen was divided into a feeding (ca. 43m 2) and lying area (ca. 58m 2). The feeding area contained the electronic sow feeding station and had a fully slatted floor. The lying area included six resting areas (each ca. 5.4m 2) with solid floors. On three sides the resting areas were separated by the house wall and 100cm high, closed plastic walls, respectively. The corridor of the lying area had a fully slatted floor. In the gilt pen, water was ad libitum accessible via two water troughs and one nipple drinker. Regardless of their feeding demand, all gilts had free access to the feeding station on the first three days in order to accustom the animals to the electronic feeding system.
28 2.2. Behavioural observation
PIG-28 Immediately after weaning, 4 rearing pens with female breeding piglets (LW and LW x DL) were filmed for 48 hours using the HeiTel Player software program (HeiTel Digital Video GmbH, Kiel, Germany). For the behavioural observation, 10 focus piglets per pen were randomly chosen and individually marked on their backs. The recording of the weaned pigs started between 15:00 and 16:45 h. The entire observation area was 3.20m x 2.10m, excluding the area around the wet feeder. In total, the agonistic behaviour of 884 piglets with a total of 32,332 fights was recorded.
PIG-68 A second observation was carried out 40 days after weaning (growing pigs = PIG-68). The same 40 focus pigs which had already been observed at weaning were re-marked, re-mixed and re-allocated to the 4 rearing pens. At the second mixing, a new group was created in equal shares from 4 pens which were previously housed at weaning. Thus, one out of four from the new group and also 2 or 3 of the marked pigs already knew each other. After mixing, the pigs were videotaped for 48 hours. The recording (PIG-68) started between 11:40 and 15:25 h. Due to problems with the technical supply only the data from 351 growing pigs with a total of 3,273 agonistic interactions were recorded. 39 of these pigs had only observations for the first 29 hours. For 40, 20 and 32 pigs, the last six, four and three hours, respectively, were missing. To include these animals in the 48-hour observation study, the incomplete data of these pigs were extrapolated based on the observation from pigs with a complete data set. Previous analysis indicated a characteristic course of the number of fights per hour after mixing (Stukenborg et al., 2010). For example, the mean time course showed that 88 percent of the fights were carried out within the first 44 hours. Due to a breakdown after 44 hours the values for a 48-hour observation were obtained by dividing the number of fights for each pig after 44 hours by 0.88. Thus, for example, pigs with 22 fights after 44 hours were assigned 25 fights after 48 hours. Hence, the extrapolation method described the individual level of each pig by considering the mean time course of all animals.
GILTS A third 48-hour observation of the animals was carried out in the gestation house. Before the gilts were moved into the pen, each animal was individually marked. The recording period started immediately after housing the gilts, between 11:35 and 16:35 h. The behaviour of all
29 animals was observed in the whole pen area. Overall, the observations from 389 including 1,536 agonistic interactions were considered for the estimation of heritabilities. Based on the technical problems during data recording, the amounts of behaviour data obtained was smaller than planned for the study particularly with regard to tracking the pigs’ behaviour across the three age groups. Overall, 239 pigs were observed as weaned piglet and growing pig. 56 animals were observed as weaned piglet and gilt and 41 pigs were observed as growing pig and gilt.
Analysing agonistic interactions The agonistic behaviour of the marked pigs was analysed with the aid of the videotapes. The growing pigs and gilts were evaluated by one person. Four observers recorded the agonistic interactions for the weaned piglets. Prior to visual assessment, the observers were trained in the definition and identification of agonistic behaviour. Characteristically distinctive fighting sequences were studied to standardise the ratings of the fights. Finally, they were tested with the same training videotape. An agonistic interaction (fight) was defined as a fight or a displacement with physical contact initiated by one individual and featuring aggressive behavioural elements, followed by any form of submissive behaviour performed by the opponent (Langbein and Puppe, 2004). Agonistic behaviour was recorded when the fights took longer than 1s. If there was an intervening period of more than 8s, a new fight was deemed to have been started (Puppe, 1998). In accordance with Puppe (1998), fights were displayed in the form of bodily attacks such as ‘head to head knocks’ and ‘head to body knocks’, ‘parallel/inverse parallel pressings’, ‘bitings’ or ‘physical displacements’ which were induced by the aggressor pig. The loser of a fight was defined as the animal which showed submissive behaviour, i.e. they stopped fighting, turned away from the attack, tried to flee or were displaced from a location (Tuchscherer et al., 1998; Langbein and Puppe, 2004). The times (start time and finish time), the aggressor/receiver and the winner/loser of a fight were noted for each fight in which a marked pig was involved. If there was no clear outcome, the fight was declared as a stand-off. If an agonistic interaction moved out of the observation area (PIG-28 and PIG-68) for more than 1 minute, the fight was terminated and the outcome was defined as “unclear”. Based on the analysed videotapes ten agonistic behavioural traits were defined and calculated for each observed pig (Table 1).
30 Table 1: Definitions of the agonistic behavioural traits Behavioural trait Unit Definition Fights no. of Number of all agonistic interactions per pig in which the pig was involved Time Fights seconds (s) The total time a marked pig was involved in agonistic interactions during the 48-hour observation period Aggressive Fights no. of The number of fights per pig in which the pig initiated the agonistic interaction Time Aggressive Fights seconds (s) The total time a marked pig was involved in self-initiated agonistic interactions during the 48-hour observation period Aggressions Received no. of The number of fights per pig in which the pig was subjected to (not the initiator of) the agonistic interaction Time Aggressions Received seconds (s) The total time a pig was subjected to agonistic interactions during the 48-hour observation period Fights Won no. of The number of fights per pig in which the pig won the agonistic interaction Time Fights Won seconds (s) The total time a marked pig was involved in won agonistic interactions during the 48-hour observation period Fights Lost no. of The number of fights per pig in which the pig lost the agonistic interaction Time Fights Lost seconds (s) The total time a pig was involved in lost agonistic interactions during the 48-hour observation period
2.3 Statistical analysis
Data analysis was performed using the SAS statistical software package (SAS Institute Inc., 2005). The distribution of the behavioural traits was analysed by using the UNIVARIATE procedure in SAS (2005). All behavioural traits had obviously skewed distributions. The original data showed a high level of skewness from 1.13 to 6.49 for all age groups. The kurtosis ranked from 0.99 to 58.04. Due to the non-normal distribution, the medians were used (Proc MEANS; SAS 2005) for descriptive statistics of the original data (Table 2).
31 Table 2: Median, skewness and kurtosis 1 of original behavioural traits for three age groups (weaned piglets = PIG-28, growing pigs = PIG-68, gilts = GILTS) Original data Median Skewness Kurtosis PIG-28 (n = 884) Fights per pig 39.0 1.22 2.66 Fight time per pig 2470 1.22 2.42 Aggressive fights per pig 13.5 2.12 8.11 Time in aggressive fights per pig 782 1.54 3.78 Aggressions received per pig 15.0 1.15 2.16 Time in aggressions received per pig 712 2.00 7.59 Fights won per pig 5.0 2.38 9.22 Time in fights won per pig 216 2.06 5.58 Fights lost per pig 5.5 1.71 5.12 Time in fights lost per pig 249 1.88 5.36 PIG-68 (n = 351) Fights per pig 8.0 2.11 6.33 Fight time per pig 106 4.30 24.61 Aggressive fights per pig 2.0 3.14 14.18 Time in aggressive fights per pig 20 4.70 31.84 Aggressions received per pig 4.0 1.41 2.20 Time in aggressions received per pig 50 5.75 49.34 Fights won per pig 0 3.69 22.54 Time in fights won per pig 0 4.92 29.89 Fights lost per pig 2.0 1.13 0.99 Time in fights lost per pig 27 6.49 58.04 GILTS (n = 389) Fights per pig 6.0 1.25 2.05 Fight time per pig 299 2.59 10.42 Aggressive fights per pig 2.0 1.92 5.68 Time in aggressive fights per pig 94 2.38 6.76 Aggressions received per pig 3.0 1.16 1.59 Time in aggressions received per pig 105 2.95 11.57 Fights won per pig 1.0 2.38 9.44 Time in fights won per pig 34 2.82 9.38 Fights lost per pig 2.0 1.18 1.48 Time in fights lost per pig 79 2.59 8.65 1Kurtosis of normal distribution = 0
In order to normalise the data, the skewness and also kurtosis were minimised by a square root (for PIG-28) or log transformation Y = log e (1+observation) (for PIG-68 and Gilts), respectively (Table 3). Due to the transformation, the data were approximated to normal distribution, with a skewness of -0.95 to 1.13 and a kurtosis of -1.53 to 0.79. Corresponding to Turner et al. (2009), a visual inspection of the distribution plots showed substantial improvements towards normality. All further analyses were based on the transformed data.
32 Table 3: Median, mean, standard deviation (S.D.), skewness and kurtosis 1 of transformed behavioural traits for three age groups (weaned piglets = PIG-28, growing pigs = PIG-68, gilts = GILTS) Transformed data Median Mean S.D. Skewness Kurtosis PIG-28 (n = 884) Fights per pig 6.2 6.4 1.8 0.24 -0.51 Fight time per pig 49.7 49.9 18.4 0.16 -0.07 Aggressive fights per pig 3.7 3.9 1.6 0.57 -0.79 Time in aggressive fights per pig 28.0 28.7 13.7 0.29 -0.14 Aggressions received per pig 3.9 3.8 1.1 0.11 -0.51 Time in aggressions received per pig 26.7 28.0 11.4 0.58 -0.66 Fights won per pig 2.2 2.4 1.5 0.50 -0.39 Time in fights won per pig 14.7 15.2 10.6 0.52 -0.12 Fights lost per pig 2.3 2.4 1.4 0.15 -0.27 Time in fights lost per pig 15.8 16.0 10.1 0.28 -0.19 PIG-68 (n = 351) Fights per pig 2.2 2.1 1.0 -0.28 -0.42 Fight time per pig 4.7 4.6 1.9 -0.48 -0.02 Aggressive fights per pig 1.1 1.2 1.0 -0.42 -0.66 Time in aggressive fights per pig 3.0 2.9 2.2 -0.00 -1.23 Aggressions received per pig 1.6 1.5 0.8 -0.25 -0.68 Time in aggressions received per pig 3.9 3.7 1.8 -0.41 -0.30 Fights won per pig 0 0.6 0.9 -1.13 -0.02 Time in fights won per pig 0 1.6 2.1 -0.86 -0.79 Fights lost per pig 1.1 1.2 0.8 -0.13 -1.00 Time in fights lost per pig 3.3 3.0 1.8 -0.36 -0.78 GILTS (n = 389) Fights per pig 1.9 1.9 0.8 -0.45 -0.31 Fight time per pig 5.7 5.2 2.0 -0.95 -0.39 Aggressive fights per pig 1.1 1.2 0.9 -0.12 -0.96 Time in aggressive fights per pig 4.6 3.9 2.5 -0.40 -1.21 Aggressions received per pig 1.4 1.4 0.7 -0.34 -0.43 Time in aggressions received per pig 4.7 4.3 2.1 -0.61 -0.41 Fights won per pig 0.7 1.0 0.9 -0.40 -1.01 Time in fights won per pig 3.6 3.2 2.6 -0.05 -1.53 Fights lost per pig 1.1 1.2 0.7 -0.23 -0.73 Time in fights lost per pig 4.4 3.9 2.2 -0.55 -0.79 1Kurtosis of normal distribution = 0
An analysis of the fixed effects was performed using the MIXED procedure in SAS (2005). The fixed effect of the observation day, the observation month and breed was tested for all behavioural traits of each age group. For model fitting, the fixed effects were added in the model stepwise. Maximum Likelihood (ML) estimation was used to test the different models. The comparisons of the models were performed using the information criteria of Hurvich and Tsai (1989) (AICC, Akaikes’ information criteria) and Schwarz (1978) (BIC, Bayesian information criteria). These values take into account the number of estimated parameters and
33 favour less complex model variants. The model with the smallest AICC and BIC values was selected without making a statement about the underlying significance. Based on the model fitting, the same model was used for all behavioural traits within an age group. The final model for the weaned piglets included the fixed effect of observation month (7 levels), the random pen effect (95 levels) and the random litter effect describing the litter in which the pig was nursed, (413 levels). The model for the growing pigs contained the observation month (8 levels), the random pen effect (37 levels) and random litter effect (146 levels). For gilts, only the observation day (8 levels) were included in the model. The random animal effect was included in the models for the heritability estimation. Heritabilities were estimated using the REML approach as implemented in the program package VCE-6 (Groeneveld et al., 2008). Due to the small sample sizes genetic correlations were not estimated. Instead, Pearson correlation coefficients were calculated between the residual of the behavioural traits.
3. Results
Behavioural performance Table 2 and Table 3 show the differences in agonistic behaviour between the three age groups. The most fighting interactions could be observed for PIG-28. All traits were obviously higher for the weaned piglets than for the older age groups. Growing pigs were more involved in fights than gilts but the median fight time for GILTS was considerably longer than for PIG-68. A more detailed description of agonistic behaviour from weaned and growing pigs is described in Stukenborg et al. (2010).
Fixed and random litter and pen effects The observation month (PIG-28, PIG-68) and observation day (GILTS), respectively, had significant influence (p < 0.05) on the behavioural traits observed. However, seasonal influences were not ascertained. For each behavioural trait, the estimated values of litter and pen variance are presented in Tables 4 (PIG-28) and 5 (PIG-68). For weaned piglets, up to 23% (Aggressions Received) of the whole variance was explained by the random pen effect. In contrast, for the growing pigs, the influence of the pen effect was lower, ranging from between 0 and 15% (Fights Lost). By comparison, the litter effect had more influence on the agonistic behaviour of weaned piglets than growing pigs. For PIG-28, the litter effect explained between 5% (Time Fights Lost) and 22% (Fights Won) of variance. For PIG-68,
34 seven traits were slightly influenced by the random litter effect, with the highest ratio of variance from 9% (Time Fights).
2 2 2 2 Table 4: Additive ( σa ), litter ( σli ), pen ( σpe ) and error ( σe ) variance with heritability estimates of the analysed behavioural traits for weaned piglets (n = 884) (standard error in parenthesis) 2 2 2 2 2 σa σli σpe σe h Fights 1) 00.31 00.45 00.33 00 1.84 0.11 (0.10) Time Fights 44.79 33.61 15.93 224.44 0.14 (0.10) Aggressive Fights 00.42 00.44 00.08 00 1.66 0.16 (0.11) Time Aggressive Fights 35.82 13.48 04.17 126.7 0.20 (0.10) Aggressions Received 00.07 00.08 00.27 00 0.75 0.06 (0.11) Time Aggressions Received 09.07 10.06 09.85 094.38 0.07 (0.09) Fights Won 00.07 00.43 00.21 00 1.24 0.04 (0.05) Time Fights Won 00.00 17.84 10.11 079.58 0.00 (0.00) Fights Lost 00.00 00.30 00.21 00 1.25 0.00 (0.00) Time Fights Lost 18.51 04.77 11.00 063.92 0.19 (0.17) 1) Descriptions for the behavioural traits are given in Table 1.
Heritabilities The heritabilities of the observed behavioural traits for weaned piglets are shown in Table 4. The highest heritabilities were estimated for the fight time of lost and aggressive fights (h 2 = 0.19 and 0.20, No additive variance component was found for the traits Time Fights Won and Fights Lost respectively. For the growing pigs, the heritabilities were generally higher in comparison to PIG-28, except for the Time Aggressive Fights and Time Fights Lost (Table 5).
2 2 2 2 Table 5: Additive ( σa ), litter ( σli ), pen ( σpe ) and error ( σe ) variance with heritability estimates of the analysed behavioural traits for growing pigs (n = 351) (standard error in parenthesis) 2 2 2 2 2 σa σli σpe σe h Fights 1) 0.26 0.01 0.06 00.64 0.26 (0.16) Time Fights 0.55 0.33 0.09 02.73 0.15 (0.17) Aggressive Fights 0.26 0.02 0.00 00.66 0.28 (0.18) Time Aggressive Fights 0.62 0.29 0.01 03.95 0.13 (0.16) Aggressions Received 0.14 0.00 0.07 00.43 0.22 (0.11) Time Aggressions Received 0.49 0.25 0.12 02.43 0.15 (0.14) Fights Won 0.31 0.07 0.00 00.45 0.37 (0.21) Time Fights Won 1.03 0.31 0.00 03.23 0.23 (0.16) Fights Lost 0.06 0.00 0.08 00.39 0.12 (0.10) Time Fights Lost 0.62 0.00 0.18 02.43 0.19 (0.13) 1) Descriptions for the behavioural traits are given in Table 1.
35 The heritability ranged from 0.12 (Fights Lost) up to 0.37 (Fights Won). The traits describing the number of interactions had considerably higher heritabilities than their corresponding “fight time” trait, except for “lost” traits. The estimated heritabilities for gilts were considerably smaller than for the younger age groups (h 2 ≤ 0.03). Higher heritabilities were only calculated for the traits concerning the receiving and loosing agonistic interactions (h 2 Aggressions Received = 0.42; h 2 Time Aggressions Received = 0.15; h 2 Fights Lost = 0.13; h 2 Time Fights Lost = 0.09).
Relationship between behavioural traits within age groups The correlations between the behavioural traits within the three age classes are shown in Table 6. Behaviour traits containing the number of fights were highly correlated with the corresponding trait describing the fight time. Therefore, only the traits concerning the number of fights were illustrated.
Table 6: Pearson correlation coefficients between agonistic behavioural traits within an age group (weaned piglets = PIG-28, growing pigs = PIG-68, gilts = GILTS) Aggressions Aggressive fights Fights Won Fights Lost Received PIG-28 Fights 0.86* 0.67* 0.69* 0.43* Aggressive Fights 0.37* 0.70* 0.40* Aggressions Received 0.30* 0.42* Fights Won 0.04* PIG-68 Fights 0.84* 0.85* 0.73* 0.58* Aggressive Fights 0.55* 0.79* 0.38* Aggressions Received 0.54* 0.57* Fights Won 0.19* GILTS Fights 0.84* 0.82* 0.76* 0.62* Aggressive Fights 0.42* 0.81* 0.38* Aggressions Received 0.47* 0.67* Fights Won 0.15* 1) Descriptions for the behavioural traits are given in Table 1. * p < 0.05
The calculated correlation coefficients were all significant and showed the same tendencies in all age groups. In every age class, Aggressive Fights were considerably more highly correlated with Fights Won than Fights Lost. Conversely, Aggressions Received were more 36 highly correlated with Fights Lost than Fights Won. Only small relationships were found between the number of fights won and the number of fights lost.
Relationship of agonistic behaviour between different age groups Table 7 shows the correlation of the same trait between the different age groups. The correlations between weaned piglets and growing pigs, ranging from -0.01 and 0.20, were rather small. Significant relationships were found for aggressive traits (Aggressive Fights r = 0.15; Time Aggressive Fights r = 0.14) and Time Fights Won (r = 0.20). Correlations between PIG-28 and GILTS were at a similar level between PIG-28 and PIG-68. However, significant relationships were not found. Moderate correlations were calculated between growing pigs and gilts, except for Fights Lost and Time Fights Lost. The most significant relationship was found for the number of fights (r = 0.47). Traits concerning the number of interactions had a significantly and obviously higher correlation coefficient than the corresponding “fight time” trait.
Table 7: Pearson correlation coefficients of agonistic behavioural traits between different age groups (weaned piglets = PIG-28, growing pigs = PIG-68, gilts = GILTS) PIG-28 – PIG68 PIG-28 – GILTS PIG-68 – GILTS (n = 239) (n = 56) (n = 41) Fights 1) -0.12 * -0.09 0.47* Time Fights -0.11 * -0.06 0.30 * Aggressive Fights -0.14* -0.07 0.37* Time Aggressive Fights -0.13* -0.04 0.22 * Aggressions Received -0.04 * -0.19 0.38* Time Aggressions Received -0.00 * -0.06 0.30 * Fights Won -0.12 * -0.02 0.31* Time Fights Won -0.20* -0.13 0.17 * Fights Lost -0.09 * -0.01 0.07 * Time Fights Lost -0.01 * -0.05 0.07 * 1) Descriptions for the behavioural traits are given in Table 1. * p < 0.05
4. Discussion
Influence of litter and pen effect Both pen and litter effect showed considerable influence on agonistic behaviour, especially for weaned piglets. The litter effect described both maternal genetic and maternal environmental effects (Roehe et al., 2009). For pigs, the litter effect is usually known as a random effect in models describing traits such as feed conversion, weight gain or piglets 37 survival (Johnson et al., 1999; Kuhlers et al., 2003; Roehe et al., 2009). With relation to behavioural influences, previous findings in rodents has shown that an offspring’s behavioural performance is influenced by the mother’s behaviour. For example, the offspring of rat mothers which show high levels of grooming and licking become less fearful animals than the offspring of mothers which show a low level on grooming and licking in the rearing period (Sapolsky, 2004; Weaver et al., 2004). Champagne (2008) suggested that such epigenetic effects can be transmitted across generations. Considering this, it can be assumed that pre- weaning experience for piglets, illustrated by the litter effect, can also influence the agonistic behaviour directly after weaning and probably beyond further stages of life. However, in the present study, the influence of the litter effect seemed to decrease over the time since the litter had a considerably higher influence on the behaviour of weaned piglets than of growing pigs. The pen effect considered different levels of aggressive behaviour between the pens. These pen differences probably resulted from the fact that certain combinations of types of animals result in greater aggression in a pen as a whole (D'Eath et al., 2009). Since the piglets were sorted by body size in order to obtain uniform animal groups, the pen effect in the present study was primarily caused through weight differences between the rearing pens. Thus, heavier pigs in pens fought more than smaller animals in pens. This would be in accordance with Andersen et al. (2000), who indicated that larger pigs spent significantly more time in fighting than smaller pigs.
Heritabilities The present investigation showed a substantial change in the heritability of agonistic behavioural traits across the three age groups. Directly after weaning, the agonistic behaviour seemed to be rather less heritable. The highest heritabilities were calculated for PIG-68. For gilts, a substantial additive genetic effect was only revealed for four behaviour traits. The small heritabilities for weaned piglets were probably an indication of the different fighting motivations between younger and older pigs. Considering that agonistic behaviour from weaned piglets is also influenced by playfulness (see also further explanations concerning the correlations between the age groups), a clear statement concerning the heritability of young piglets’ agonistic behaviour is not possible. For growing pigs, traits measuring the number of fights, except Fights Lost, had considerably higher heritability estimates than the corresponding fight time traits. These differences might be caused by the imprecise determination of fight time traits. Especially growing pigs had a high number of short fights whereby even small deviations in the precise fight time lead to considerably vague time
38 allocations. Considering this, the number of fights traits seemed to be more appropriate for a behavioural description than the analogous fight time. Regarding previous investigations in agonistic behaviour for growing pigs, Turner et al. (2008; 2009) estimated heritabilities for the duration in reciprocal aggression (h 2 = 0.46 and 0.43) and the durations in either delivery (h 2 = 0.37 and 0.31) or receipt of non-reciprocal aggression (h 2 = 0.17 and 0.08). By comparison, Time Fights in the present study, corresponding to duration in reciprocal aggression, had a clearly smaller heritability (h 2 = 0.15). However, differences between heritability estimates are probably based on differences in the experimental design. Turner et al. (2008; 2009) analysed agonistic behaviour for 24 hours in a pen with 15 instead of 29 pigs (as performed in the present study). Therefore, the behaviours of each pig were better observable. In addition, the traits analysed from Turner et al. (2008; 2009) had a partly different definition than the present traits. For example, they distinguished whether a pig retaliated or stayed passive after a delivery of aggression. Submissive manners seemed to be heritable for gilts, especially the receiving and the losing traits. Løvendahl et al. (2005) described intermediate heritabilities for performed aggression traits (h 2 = 0.17 to 0.24). Conversely, they estimated low heritabilities for received aggression (h 2 = 0.04 to 0.06). They concluded that aggressiveness was a heritable trait of the aggressor but less so a trait of the receiver. However, the lack of further moderate estimates in the present study was probably caused by housing effects which influenced the behavioural judgment for gilts. Indeed, every agonistic interaction in the gilt pen was recorded and analysed, but due to the relatively high space allowance and the visual barriers provided the gilts could keep away or escape from fighting conflicts (Arey and Edwards, 1998; Weng et al., 1998; Spoolder et al., 2009). Since the gilts could avoid agonistic interactions, it might be that a clear estimation of the sows’ agonistic potential was not always stated. In this respect, Løvendahl et al. (2005) tried to provoke fighting behaviour by keeping 5 to 10 sows in a testing arena for 30 minutes. Their results indicated that induced fighting activities could lead to other findings than those of the present study. Probably, an intentionally applied temperament test can better reflect the agonistic potential of gilts.
Correlations within and between age groups For all three age groups the correlations between Aggressive Fights/Aggressions Received and Fights Won/Fights Lost, respectively, indicated that an aggressor pig which initiated a fight more often was also more often the winner. Conversely, submissive animals, which were mainly the losing pigs at the end of an agonistic interaction, were mostly attacked at the start
39 of a fight. The correlation between the age classes showed that the behaviour of weaned piglets was only weakly correlated with the behaviour of the older age groups. By contrast, intermediate correlations were calculated between growing pigs and gilts. The differences between PIG-28 on one the one hand and PIG-68 and gilts on the other hand could be an indication for two different motivations for fighting. The fighting behaviour from the older age groups seemed to be partly comparable and tendential, more aggressive growing pigs also became more aggressive sows. However, it can not be assumed that weaned piglets with a high level of agonistic behaviour were also more aggressive growing pigs and gilts, respectively. Regarding weaned piglets’ agonistic behaviour is probably related to playful behaviour. In different studies (Dobao et al., 1985; Chaloupková et al., 2007) playful behaviour of weaned piglets has been defined as similar to agonistic behaviour. Donaldson et al. (2002) remarked that social play can also occur during serious aggression. According to Silerová et al., (2010) there seems to be no clear separation between fighting and playful behaviour in the current literature. They suggested that playful and fighting behaviour formed one continuum. Based on this, it can be supposed that the applied behavioural traits for weaned piglets also included a large non-agonistic component. In contrast, a relationship between playful behavioural patterns and agonistic behaviour for growing pigs and gilts has never been described in previous studies. In these age groups, agonistic interactions especially occur for feeding competitions (Csermely and Wood-Gush, 1987) or in order to establish a new dominance hierarchy (Arey, 1999). Due to the different fighting motivations, the agonistic behaviour of weaned piglets cannot be compared with that of the older age groups.
5. Conclusion
Under commercial farm conditions, the highest heritabilities for agonistic behaviour will be estimated for growing pigs. A more precise behavioural description can be made more via traits considering the number of agonistic behaviours than via their corresponding fight time trait. The only slight correlations between weaned piglets and older age groups are probably caused by two different fighting motivations. The agonistic behaviour of weaned piglets is probably related to playfulness while older pigs seem to establish a new dominance hierarchy. Finally, the results indicate a considerable influence of the litter effect and therefore of a maternal effect on agonistic behaviour. In further analysis this effect should be specified over the whole ontogenesis of a pig.
40
References
Andersen, I.L., Andenæs, H., Bøe, K.E., Jensen, P., Bakken, M., 2000. The effects of weight asymmetry and resource distribution on aggression in groups of unacquainted pigs. Applied Animal Behaviour Science 68, 107-120. Arey, D.S., 1999. Time course for the formation and disruption of social organisation in group-housed sows. Applied Animal Behaviour Science 62, 199-207. Arey, D.S., Edwards, S.A., 1998. Factors influencing aggression between sows after mixing and the consequences for welfare and production. Livestock Production Science 56, 61-70. Chaloupková, H., Illmann, G., Bartoš, L., Špinka, M., 2007. The effect of pre-weaning housing on the play and agonistic behaviour of domestic pigs. Applied Animal Behaviour Science 103, 25-34. Champagne, F.A., 2008. Epigenetic mechanisms and the transgenerational effects of maternal care. Frontiers in Neuroendocrinology 29, 386-397. Csermely, D., Wood-Gush, D.G.M., 1987. Aggressive Behaviour of Grouped Sows in Different Contexts. Applied Animal Behaviour Science 17, 368-369. D'Eath, R.B., Roehe, R., Turner, S.P., Ison, S.H., Farish, M., Jack, M.C., Lawrence, A.B., 2009. Genetics of animal temperament: aggressive behaviour at mixing is genetically associated with the response to handling in pigs. Animal 3, 1544-1554. Dobao, M.T., Rodrigañez, J., Silio, L., 1985. Choice of companions in social play in piglets. Applied Animal Behaviour Science 13, 259-266. Donaldson, T.M., Newberry, R.C., Špinka, M., Cloutier, S., 2002. Effects of early play experience on play behaviour of piglets after weaning. Applied Animal Behaviour Science 79, 221-231. Ewbank, R., 1976. Social hierarchy in suckling and fattening pigs: A review. Livestock Production Science 3, 363-372. Groeneveld, E., Kovac, M., Mielenz, N., 2008. VCE User’s Guide and Reference Manual Version 6.0. Institute of Farm Animal Genetics, Friedrich Loeffler Institute (FLI), Mariensee, Germany. Hurvich, C.M., Tsai, C.L., 1989. Regression and Time-Series Model Selection in Small Samples. Biometrika 76, 297-307.
41 Johnson, Z.B., Chewning, J.J., Nugent, R.A., 3rd, 1999. Genetic parameters for production traits and measures of residual feed intake in large white swine. J. Anim Sci. 77, 1679- 1685. Kuhlers, D.L., Nadarajah, K., Jungst, S.B., Anderson, B.L., Gamble, B.E., 2003. Genetic selection for lean feed conversion in a closed line of Duroc pigs. Livestock Production Science 84, 75-82. Langbein, J., Puppe, B., 2004. Analysing dominance relationships by sociometric methods - a plea for a more standardised and precise approach in farm animals. Applied Animal Behaviour Science 87, 293-315. Løvendahl, P., Damgaard, L.H., Nielsen, B.L., Thodberg, K., Su, G., Rydhmer, L., 2005. Aggressive behaviour of sows at mixing and maternal behaviour are heritable and genetically correlated traits. Livestock Production Science 93, 73-85. Puppe, B., 1998. Effects of familiarity and relatedness on agonistic pair relationships in newly mixed domestic pigs. Applied Animal Behaviour Science 58, 233-239. Roehe, R., Shrestha, N.P., Mekkawy, W., Baxter, E.M., Knap, P.W., Smurthwaite, K.M., Jarvis, S., Lawrence, A.B., Edwards, S.A., 2009. Genetic analyses of piglet survival and individual birth weight on first generation data of a selection experiment for piglet survival under outdoor conditions. Livestock Science 121, 173-181. Sapolsky, R.M., 2004. Mothering style and methylation. Nature Neuroscience 7, 791-792. SAS Institute Inc., 2005. SAS/STAT User's guide, Version 9.1. Schwarz, G., 1978. Estimating the dimension of a model. Annals of Statistics 6, 461-464. Šilerová, J., Špinka, M., Šárová, R., Algers, B., 2010. Playing and fighting by piglets around weaning on farms, employing individual or group housing of lactating sows. Applied Animal Behaviour Science 124, 83-89. Spoolder, H.A.M., Geudeke, M.J., Van der Peet-Schwering, C.M.C., Soede, N.M., 2009. Group housing of sows in early pregnancy: A review of success and risk factors. Livestock Science 125, 1-14. Stookey, J.M., Gonyou, H.W., 1994. The effects of regrouping on behavioral and production parameters in finishing swine. Journal of Animal Science 72, 2804-2811. Stukenborg, A., Traulsen, I., Puppe, B., Presuhn, U., Krieter, J., 2010. Agonistic behaviour after mixing in pigs under commercial farm conditions. Applied Animal Behaviour Science (2010), doi:10.1016/j.applanim.2010.10.004.
42 Tan, S.S.L., Shackleton, D.M., Beames, R.M., 1991. The effect of mixing unfamiliar individuals on the growth and production of finishing pigs. Animal Production 52, 201-206. Tuchscherer, M., Manteuffel, G., 2000. The effect of psycho stress on the immune system. Another reason for pursuing animal welfare (Review). Archiv für Tierzucht - Archives of Animal Breeding 43, 547-560. Tuchscherer, M., Puppe, B., Tuchscherer, A., Kanitz, E., 1998. Effects of social status after mixing on immune, metabolic, and endocrine responses in pigs. Physiology & Behavior 64, 353-360. Turner, S.P., Roehe, R., D'Eath, R.B., Ison, S.H., Farish, M., Jack, M.C., Lundeheim, N., Rydhmer, L., Lawrence, A.B., 2009. Genetic validation of postmixing skin injuries in pigs as an indicator of aggressiveness and the relationship with injuries under more stable social conditions. Journal of Animal Science 87, 3076-3082. Turner, S.P., Roehe, R., Mekkawy, W., Farnworth, M.J., Knap, P.W., Lawrence, A.B., 2008. Bayesian analysis of genetic associations of skin lesions and behavioural traits to identify genetic components of individual aggressiveness in pigs. Behavior Genetics 38, 67-75. van Erp-van der Kooij, E., Kuijpers, A.H., van Eerdenburg, F.J.C.M., Tielen, M.J.M., 2003. Coping characteristics and performance in fattening pigs. Livestock Production Science 84, 31-38. Weaver, I.C.G., Cervoni, N., Champagne, F.A., D'Alessio, A.C., Sharma, S., Seckl, J.R., Dymov, S., Szyf, M., Meaney, M.J., 2004. Epigenetic programming by maternal behavior. Nature Neuroscience 7, 847-854. Weng, R.C., Edwards, S.A., English, P.R., 1998. Behaviour, social interactions and lesion scores of group-housed sows in relation to floor space allowance. Applied Animal Behaviour Science 59, 307-316.
43
44 CHAPTER THREE
Relationship between agonistic behaviour, growth and reproductive performance in pigs
Andreas Stukenborg 1, Imke Traulsen 1; Eckhard Stamer 2, Birger Puppe 3, Ulrich Presuhn 4, Joachim Krieter 1
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, D-24098 Kiel, Germany
2TiDa Tier und Daten GmbH, D-24259 Westensee/Brux, Germany
3Leibniz Institute for Farm Animal Biology (FBN), D-18196 Dummerstorf, Germany
4farm concepts GmbH & Co. KG, D-23812 Wahlstedt, Germany
45 Abstract
The aim of the present study was to analyse the relationship between agonistic pig behaviour and common performance traits (reproduction and growth performance) in pig breeding. Agonistic behaviour was observed directly after mixing pigs over a 48-hour period in three different age groups: after weaning (PIG-28), 40 days later (growing pigs, PIG-68) and immediately after mixing gilts (GILTS). The agonistic interactions were analysed by noting the times (start and finish), the aggressor/receiver and the winner/loser of a fight. Based on these results, each pig was given information regarding ten agonistic behavioural traits. Live daily gain and backfat depth were measured in an own performance test seven weeks before the gilts were housed in the pregnancy area. In order to consider the reproductive performance of a sow, the number of total born piglets, the number of piglets born alive and the number of stillborn piglets were noted for each litter. Except for the number of initiated fights and the time spent in initiated fights no significant relationships were found between growth performance traits and the agonistic behaviour of weaned piglets. Concerning growing pigs, five traits were significantly correlated with backfat depth, with correlation coefficients ranging from r = 0.20 up to r = 0.31. No clear relationship was found between the behaviour of PIG-68 and live daily gain. However, a positive correlation was estimated between live daily gain and the behaviour of gilts (r = 0.08 up to r = 0.26). Except for the number of aggressions received and the time spent in received aggressions, a clear relationship between gilts’ behaviour and backfat depth was not estimated. No significant correlations were found between the agonistic behaviour of newly mixed gilts and their later reproductive performance. In conclusion, the results indicate that the agonistic behaviour of weaned piglets was not related to their subsequent growth performance. For growing pigs, backfat depth was positive correlated with agonistic behaviour. Gilts with a higher live daily gain seemed to be more involved in agonistic interactions. However, the behavioural ascertainment for gilts was probably influenced by the considerable space allowance. In further studies the behaviour of gilts should be observed in induced stress situations, which can probably better reflect their agonistic potential.
Keywords: pig, agonistic behaviour, correlation, growth and reproductive performance, age groups
46 1. Introduction
Agonistic behaviour in pigs can usually be observed when unacquainted pigs are newly mixed. Due to fighting interactions a clear dominance hierarchy will be established (Ewbank, 1976). In pig husbandry, pigs are commonly mixed three times, directly after weaning, after leaving the rearing pens and when breeding sows are kept in the pregnancy area. Frequently, involvement in agonistic interactions leads to skin injuries. The pigs become stressed (Björk, 1989), which in turn can cause immunosuppressive effects (Tuchscherer and Manteuffel, 2000). Consequently, agonistic behaviour results in decreased welfare and can negatively affect the economy of pig production, for example by reduced weight gain (Tan et al., 1991; Stookey and Gonyou, 1994). Due to the unwanted concomitant effects, a decrease in aggressiveness in pig breeding seems to be useful. D’Eath et al. (2009) stated that agonistic behaviour is moderately heritable and could be reduced by genetic selection. However, there is a lack of studies which deal with the relationships between agonistic behaviour and common performance traits (reproduction and growth performance) in pig breeding. Cassady (2007) found a positive phenotypic correlation between lean growth and aggressiveness. By contrast, Turner et al. (2006) found neither a phenotypic nor genotypic correlation between lesion score, indicating aggressiveness, and growth rate or backfat depth. Concerning fertility traits, Kranendonk et al. (2007) and Jarvis et al. (2006) assumed that there was no relationship between social ranking and litter size or percentage of live born piglets. Nevertheless, it is still uncertain whether the current selection practice in pig breeding has an influence on the agonistic behaviour of pigs. Conversely, it remains unexplained whether a greater consideration of agonistic behaviour in pig breeding would have negative effects on common performance traits. Therefore, the objective of the present paper was to analyse the relationships between agonistic behaviour, observed for three common mixing times and growth rate gain and backfat depth, respectively. In addition, correlations were estimated between reproductive performance and the agonistic behavioural traits of gilts.
47 2. Material and Methods
2.1. Animals and Housing
The data collection was done from October 2007 to April 2009 on a closed herd sow farm of the German breeding company ‘Hülsenberger Zuchtschweine’. The herd composed of pure- bred Large-White (LW) sows and also crosses between Large-White and German Landrace (LW x DL). For the present study female pigs from pure bred litters and crosses from LW x DL were considered. The sows were managed in a one-week rhythm with a 28-day lactating period. About one week before the calculated farrowing date, the sows of a farrowing batch were removed from the pregnancy area and washed. Subsequently, the averagely 36 sows were kept in the farrowing area. All farrowing pens had the same dimensions of 1.7m x 2.6m. Every Wednesday, piglets from one farrowing batch were weaned and moved to the rearing house where they were sorted by gender and body size. The pig behaviour was observed (see Section 2.2) at three different stages of life: weaned piglets (PIG-28), growing pigs (PIG-68) and gilts (GILTS). The housing conditions during the three observation times are described below.
PIG-28/PIG-68 Both weaned piglets and growing pigs were observed in the rearing pens. After weaning, 29 piglets on average were housed in a rearing pen. The dimensions of the pens were 5.10m x 2.10m. The pens had a fully slatted floor and were separated by 60cm high, closed partitions on the ground and two rods across. During the rearing period the piglets were fed ad libitum with a wet feeder. Water was non-stop accessible on nipple drinkers.
GILTS After the rearing period the pigs were taken to a fattening unit where the own performance test was carried out. In the fattening unit, 12 pigs on average were housed in a pen. The pigs were fed with a liquid feeding system three times a day. The third observation was carried out when the gilts were brought back to the sow farm, where the gilts were housed as a single group in the pregnancy area. 48 gilts on average were housed in the group pen, with dimensions of ca. 15.5m x 6.5m. Before the gilts were brought together. All of them were washed and were given an earmark for the feeding station. The pen consisted of a feeding (ca. 43m 2) and a lying area (ca. 58m2). The feeding area had a fully slatted floor and contained the electronic sow feeding station. The lying area contained six resting areas (each ca. 5.4m 2)
48 with a solid floor. The areas were separated on three sides by the house wall and 100cm high, closed plastic walls, respectively. The corridor of the lying area had a fully slatted floor. Water was accessible ad libitum via two water troughs and one nipple drinker. In order to accustom the animals the electronic feeding system, all of them had free access to the feeding station for the first three days.
2.2. Behavioural observation
PIG-28 Four rearing pens with female breeding piglets were filmed directly after weaning using the HeiTel Player software program (HeiTel Digital Video GmbH, Kiel, Germany). The behavioural observations were made for 10 focus piglets per pen, which were randomly chosen and individually marked on their backs. The recording of the weaned piglets started between 15:00 and 16:45 h. The entire observation area measured 3.20m x 2.10m and excluded the area around the wet feeder.
PIG-68 A second observation took place 40 days after weaning (growing pigs = PIG-68). A new group was created in equal shares from 4 pens which were previously housed at weaning. The same 40 focus pigs which had already been observed at weaning were remarked, remixed and reallocated to the 4 rearing pens. Thus, one out of four from the new group and also 2 or 3 from the marked pigs were already acquainted. After mixing, the pigs were videotaped for 48 hours. The recording (PIG-68) started between 11:40 and 15:25 h. Due to problems with the technical supply, 21 from 97 growing pigs were only observed for the first 30 (one animal), 42 or 44 hours. To include these animals in the 48 hour observation study, the incomplete data of these pigs were extrapolated based on the observation from pigs with a complete data set. Previous analysis indicated a characteristic course of the number of fights per hour after mixing (Stukenborg et al., 2010a). For example, the mean time course showed that 88 percent of the fights were carried out within the first 44 hours. Due to a breakdown after 44 hours, the values for a 48-hour observation were obtained by dividing the number of fights for each pig after 44 hours by 0.88. Thus, for example, pigs with 22 fights after 44 hours were assigned 25 fights after 48 hours. Hence, the extrapolation method described the individual level of each pig by considering the mean time course of all animals.
49 GILTS The third 48 hours observation was carried out in the gestation house. Before the gilts were moved into the pen, each animal was individually marked. The recording period started between 11:35 and 16:35 h, immediately after housing the gilts. The behaviour of all animals inside the whole pen area was observed.
Analysing agonistic interactions The agonistic behaviour of each marked pig was examined with the aid of the videotapes. A single agonistic interaction was defined as a fight or a displacement with physical contact initiated by one individual and featuring aggressive behavioural elements, followed by any form of submissive behaviour performed by the opponent (Langbein and Puppe, 2004). An agonistic interaction was recorded if the fight lasted longer than 1s. In accordance with Puppe (1998), a new fight was deemed to have been started when an intervening period took more than 8s. Fights were displayed in the form of bodily attacks such as ‘head to head knocks’ and ‘head to body knocks’, ‘parallel/inverse parallel pressings’, ‘biting’ or ‘physical displacements’ which were induced by the aggressor pig (Puppe, 1998). An animal showing submissive manners was defined as the looser of a fight. A looser stopped fighting, turned away from the attack, tried to flee or was displaced from a location etc. (Tuchscherer et al., 1998; Langbein and Puppe, 2004). The aggressor/receiver, the winner/loser of a fight and also the times (start time and finish time) were noted for each fight in which a marked pig was involved. A fight was declared as a stand-off when there was no obvious outcome. The fight was terminated and the outcome defined as “unclear” if an agonistic interaction moved out of the observation area (PIG-28 and PIG-68) for more than 1 minute. Based on the analysed videotapes ten agonistic behavioural traits were defined and calculated for each observed pig (Table 1).
50 Table 1: Definitions of the agonistic behavioural traits Behavioural trait Unit Definition Fights no. of Number of all agonistic interactions per pig in which the pig was involved Time Fights seconds (s) The total time a marked pig was involved in agonistic interactions during the 48-hour observation period Aggressive Fights no. of The number of fights per pig in which the pig initiated the agonistic interaction Time Aggressive Fights seconds (s) The total time a marked pig was involved in self-initiated agonistic interactions during the 48-hour observation period Aggressions Received no. of The number of fights per pig in which the pig was subjected to (not the initiator of) the agonistic interaction Time Aggressions Received seconds (s) The total time a pig was subjected to agonistic interactions during the 48-hour observation period Fights Won no. of The number of fights per pig in which the pig won the agonistic interaction Time Fights Won seconds (s) The total time a marked pig was involved in won agonistic interactions during the 48-hour observation period Fights Lost no. of The number of fights per pig in which the pig lost the agonistic interaction Time Fights Lost seconds (s) The total time a pig was involved in lost agonistic interactions during the 48-hour observation period
2.3. Growth and reproductive performance traits
Growth performance traits were measured during an own performance test circa seven weeks before gilts were housed in the pregnancy area. The average daily weight gain from birth till test date was defined as live daily gain. Backfat depth was measured with an ultrasonic A- Scan instrument (USN 50, Krautkramer Branson). The following three points were considered for backfat depth measurement: at the 13 th /14 th rib, 10cm behind the shoulder and 10cm in front of the ham, 6cm from the dorsal mid line (right side). The sum of backfat depth (henceforth referred to as backfat depth) was calculated by a summation of all three points. Fertility traits concerning litter size were ascertained directly after farrowing. The number of piglets born in total, the number of piglets born alive, the number of stillborn piglets and the
51 number of mummified piglets was noted for each litter. If a dead-born piglet was fully developed, it was defined as stillborn, not fully developed piglets were defined as mummified. An overview of growth and reproductive performance traits is given in Table 2.
Table 2: Mean, standard deviation (S.D.), minimum (Min) and maximum (Max) of growth (n = 199) and reproductive performance traits (n = 5710) Mean S.D. Min Max Live daily gain (g) 593.8 55.6 457 727 Backfat depth (mm) 1 31.4 6.2 17 49 Piglets born in total per litter 14.1 3.7 0 27 Piglets born alive per litter 12.8 3.5 0 24 Piglets stillborn per litter 1.2 1.7 0 16 1 Summation of three backfat depth measuring points
2.4 Statistical analysis
Statistical data analysis was performed with the SAS statistical software package (SAS Institute Inc., 2005). The distribution of the behaviour traits was analysed by using the UNIVARIATE procedure in SAS (2005). All behavioural traits had obviously skewed distributions. The best transformation of the data to obtain normality was reached by a square root (for PIG-28) or log transformation Y = log e (1+observation) (for PIG-68 and Gilts). All further analyses were based on the transformed data. An analysis of fixed effects was performed with the MIXED procedure in SAS (2005). For model fitting, the fixed effects were added in the model in a stepwise procedure. Maximum Likelihood (ML) estimation was used to test the different models. The comparisons of the models were determined with the information criteria of Hurvich and Tsai (1989) (AICC, Akaikes’s information criteria) and Schwarz (1978) (BIC, Bayesian information criteria). These values take into account the number of estimated parameters and favour less complex model variants. The model with the smallest AICC and BIC values was chosen without considering the underlying significance of fixed effects. With regard to behavioural traits, the final model included the fixed effect of observation month (PIG-28, PIG-68) and observation day (GILTS) and also the random pen effect for PIG-28 and PIG-68 (Table 3).
52 Table 3: Description of models used concerning the relationship between behavioural traits and growth performance. Number of classes in parentheses (PIG-28 / PIG-68 / GILTS) (n = 107 (PIG-28); n = 97 (PIG-68); n = 199 (GILTS)) Random Body weight Test day Season 1 pen-effect Behaviour n.c. 2 n.c. (7/6/7) (24/28/n.c.) Growth rate n.c. (5/5/10) n.c. n.c. Backfat depth cov 3 (5/5/10) n.c. n.c. 1 observation month for PIG-28 and PIG-68; observation date for GILTS 2 not considered 3 linear regression on body weight
For growth performance traits, the models included the fixed effect of the test day (live daily gain and backfat depth) and a linear regression on body weight (only for backfat depth). The final models for all three reproductive traits included the fixed effect of farrowing month, parity and the random permanent effect of the sow (Table 4).
Table 4: Description of models used concerning the relationship between behavioural traits (GILTS) and reproductive performance. Number of classes in parentheses (n = 264 for behaviour; n = 5710 for reproductive performance) Farrowing Random Parity 1 Obervation date month sow-effect Behaviour n.c. 2 n.c. (8) n.c. Reproductive (7) (35) n.c. (1840) performance 1 parity class seven = parities ≥ 7 2 not considered
Due to problems with the technical supply, the number of observed animals was considerably different between the age groups. The relationship between growth performance and behaviour of the different age groups was analysed using 107 (PIG-28), 97 (PIG-68) and 199 (GILTS) animals. Due to the small number of animals which had observations for behaviour as well reproductive performance the correlations between these traits were only estimated for GILTS (n = 264). In order to consider all possible related performances, the reproductive information from 5710 litters were included in the model. Correlation coefficients were estimated with the REML approach as implemented in the program package VCE-6 (Groeneveld et al., 2008), with including the random animal effect in the models. Genetic correlations were not estimated due to the small sample sizes. Significances of the phenotypic correlation coefficients were tested with a t-test for correlation coefficient (Kanji, 1999).
53 3. Results
Relationship between agonistic behaviour and growth performance For weaned piglets no relationship was found between live daily gain and agonistic behavioural traits (Table 5). Significant correlations were only estimated between backfat depth and Aggressive Fights (r = 0.23) and Time Aggressive Fights (r = 0.22), respectively.
Table 5: Phenotypic correlations between growth performance traits and agonistic behaviour of three different age groups Behaviour PIG-28 Behaviour PIG-68 Behaviour GILTS Live Live Live Backfat Backfat Backfat daily daily daily depth depth depth gain gain gain Fights -0.05 * -0.12 * -0.03 * 0.24* 0.14 * -0.13 * Time Fights -0.04 * -0.13 * -0.03 * 0.26* 0.20* -0.13 * Aggressive Fights -0.12 * -0.23* -0.02 * 0.25* 0.23* -0.05 * Time Aggressive Fights -0.09 * -0.22* -0.09 * 0.31* 0.26* -0.07 * Aggressions Received -0.03 * -0.10 * -0.07 * 0.20* 0.06 * -0.19* Time Aggressions Received -0.09 * -0.12 * -0.06 * 0.18 * 0.15* -0.21* Fights Won -0.04 * -0.04 * -0.15 * 0.11 * 0.14* -0.09 * Time Fights Won -0.04 * -0.14 * -0.10 * 0.10 * 0.13 * -0.10 * Fights Lost -0.04 * -0.11 * -0.17 * 0.10 * 0.08 * -0.01 * Time Fights Lost -0.01 * -0.06 * -0.20 * 0.13 * 0.22* -0.01 * * p < 0.05
Live daily gain was not significant correlated with the behaviour traits of growing pigs. Positive significant relationships were estimated between backfat and the number of fights (r = 0.24), the overall fight time duration (r = 0.26), the number of initiated fights (r = 0.25), the duration in initiated fights (r = 0.31) and the number of received aggressions (r = 0.20). For gilts, six out of ten behavioural traits were significantly positively correlated with live daily gain with the highest correlation coefficient of r = 0.26 (with Time Aggressive Fights). A significant relationship was also estimated between backfat depth and Aggressions Received (r = 0.19) and Time Aggressions Received (r = 0.21).
Relationship between agonistic behaviour and reproductive performance No significant correlations were found between reproductive performance traits and the agonistic behaviour of newly mixed gilts (Table 6). All estimated correlation coefficients
54 were relatively small, ranging from r = -0.07 to 0.04 (piglets total born per litter), r = -0.08 to 0.04 (piglets born alive per litter) and r = -0.04 to 0.03 (piglets stillborn per litter).
Table 6: Phenotypic correlations between reproductive performance traits and the agonistic behaviour of gilts Piglets total born Piglets born alive Piglets stillborn per
per litter per litter litter Fights -0.02 -0.07 -0.03 Time Fights -0.01 -0.01 -0.04 Aggressive Fights -0.04 -0.04 -0.01 Time Aggressive Fights -0.02 -0.00 -0.04 Aggressions Received -0.07 -0.08 -0.02 Time Aggressions Received -0.05 -0.04 -0.03 Fights Won -0.00 -0.00 -0.00 Time Fights Won -0.03 -0.02 -0.03 Fights Lost -0.02 -0.04 -0.03 Time Fights Lost -0.01 -0.03 -0.03
4. Discussion
Relationship between agonistic behaviour and growth performance Above the age groups no uniform results were found concerning the correlations between agonistic behavioural traits and growth performance. With the exception of the number and duration of initiated interactions, there seemed to be no relationship between the agonistic behaviour of weaned piglets and live daily gain or backfat depth. Neither a substantial correlation was found between live daily gain and the behaviour of growing pigs. However, for PIG-68, the results indicated a positive relationship between backfat depth and the involvement in agonistic interactions. A broad relationship between backfat depth and agonistic performance was not estimated for gilts. Conversely, most of the observed behavioural traits were significantly positively correlated with live daily gain. This could be an indication that heavier gilts with a higher live daily gain spend more time in agonistic interactions than lighter ones. This is in accordance with Andersen et al. (2000), who suggested that larger pigs spent significantly more time in fighting than smaller pigs. It might be that smaller, thus submissive gilts have fewer agonistic interactions because of the comparatively high space allowance in the gilt pen and the visual barriers provided which allowed the animals to flee and hide from greater, feared sows (Arey and Edwards, 1998;
55 Weng et al., 1998; Spoolder et al., 2009). The lack of correlations between growth performance traits and the behaviour of weaned piglets and growing pigs (regarding Live daily gain), respectively, might be caused by the long interval between the two observation times (behavioural vs. growth performance traits). The motivation to fight may change due to the long intervening periods. The findings of previous investigations have shown that the behaviour of weaned piglets is not correlated with the behaviour of growing pigs and gilts (Stukenborg et al., 2010b). According to this, the agonistic behaviour of weaned piglets is primarily influenced by playful manners while older pigs fight in order to establish a new dominance hierarchy. Moderate correlations were found between the behaviour of growing pigs and gilts. However, concerning the relationship with growth performance traits, the behaviour of growing pigs cannot be compared with the behaviour of gilts. This in turn could probably indicate that agonistic behaviour is far less influenced by body size than the behaviour of gilts. Turner et al. (2006) observed growing pigs and found neither a phenotypic nor genotypic correlation between lesion score, indicating aggressiveness, and growth rate or backfat depth. In their opinion, genetic selection to decrease aggression is possible without reducing the genetic improvement in growth rate or backfat depth.
Relationship between agonistic behaviour and reproductive performance All calculated correlation coefficients between agonistic behaviour and total born piglets, alive born piglets and also stillborn piglets were rather small and also non-significant. Therefore it can be stated that there was no relationship between the aggressiveness of gilts and their subsequent fertility performance. On the supposition that the social status of a sow could also be an expression of the aggressive potential, the results of Kranendonk et al. (2007) and Jarvis et al. (2006) were in accordance to the presented findings. They also found no relationship between social ranking and litter size or percentage of live born piglets. In contrast, Hoy et al. (2009) indicated a significant difference in the number of total born piglets between high- and low-ranking sows. As already mentioned, the gilts were given a great space allowance and a structured pen which allowed the animals to hide and flee. The possibility to avoid agonistic interactions could lead to an imprecise estimation of the sows’ agonistic potential. In a previous study, it was suggested that inexact behavioural ascertainment could be the reason for small heritability estimates of agonistic behavioural traits for gilts (Stukenborg et al., 2010b). The present results concerning the relationship with reproductive performance were also probably influenced by the vague behaviour determination for gilts. Therefore, in further studies, the agonistic behaviour of gilts should be
56 observed in induced stress situations. For example, Løvendahl et al. (2005) tried to provoke fighting behaviour by keeping 5 to 10 sows in a testing arena for 30 minutes. Such an intentionally applied temperament test can probably better reflect the agonistic potential of gilts (Stukenborg et al., 2010b).
5. Conclusion
The present findings suggest that there is no phenotypic relationship between the behaviour of weaned piglets and their growth performances. There seems to be a positive correlation between backfat depth and agonistic behaviour in growing pigs. In gilts, agonistic behaviour is positive correlated with their live daily gain but an involvement in agonistic interactions does not influence their subsequent reproductive performance. However, the behaviour ascertainment for gilts is probably influenced by the space allowance of the animals. In further studies, the behaviour of gilts should be observed in induced stress situations which can probably better reflect their agonistic potential.
References
Andersen, I.L., Andenæs, H., Bøe, K.E., Jensen, P., Bakken, M., 2000. The effects of weight asymmetry and resource distribution on aggression in groups of unacquainted pigs. Applied Animal Behaviour Science 68, 107-120. Arey, D.S., Edwards, S.A., 1998. Factors influencing aggression between sows after mixing and the consequences for welfare and production. Livestock Production Science 56, 61-70. Björk, A.K.K., 1989. Is Social Stress in Pigs a Detrimental Factor to Health and Growth that can be Avoided by Amperozide Treatment. Applied Animal Behaviour Science 23, 39-47. Cassady, J.P., 2007. Evidence of phenotypic relationships among behavioral characteristics of individual pigs and performance. J. Anim Sci. 85, 218-224. D'Eath, R.B., Roehe, R., Turner, S.P., Ison, S.H., Farish, M., Jack, M.C., Lawrence, A.B., 2009. Genetics of animal temperament: aggressive behaviour at mixing is genetically associated with the response to handling in pigs. Animal 3, 1544-1554.
57 Ewbank, R., 1976. Social hierarchy in suckling and fattening pigs: A review. Livestock Production Science 3, 363-372. Groeneveld, E., Kovac, M., Mielenz, N., 2008. VCE User’s Guide and Reference Manual Version 6.0. Institute of Farm Animal Genetics, Friedrich Loeffler Institute (FLI), Mariensee, Germany. Hoy, S., Bauer, J., Borberg, C., Chonsch, L., Weirich, C., 2009. Impact of rank position on fertility of sows. Livestock Science 126, 69-72. Hurvich, C.M., Tsai, C.L., 1989. Regression and Time-Series Model Selection in Small Samples. Biometrika 76, 297-307. Jarvis, S., Moinard, C., Robson, S.K., Baxter, E., Ormandy, E., Douglas, A.J., Seckl, J.R., Russell, J.A., Lawrence, A.B., 2006. Programming the offspring of the pig by prenatal social stress: Neuroendocrine activity and behaviour. Hormones and Behavior 49, 68- 80. Kanji, G.K., 1999. 100 Statistical Tests. Sage Publications Ltd London. Kranendonk, G., Van der Mheen, H., Fillerup, M., Hopster, H., 2007. Social rank of pregnant sows affects their body weight gain and behavior and performance of the offspring. Journal of Animal Science 85, 420-429. Langbein, J., Puppe, B., 2004. Analysing dominance relationships by sociometric methods - a plea for a more standardised and precise approach in farm animals. Applied Animal Behaviour Science 87, 293-315. Løvendahl, P., Damgaard, L.H., Nielsen, B.L., Thodberg, K., Su, G., Rydhmer, L., 2005. Aggressive behaviour of sows at mixing and maternal behaviour are heritable and genetically correlated traits. Livestock Production Science 93, 73-85. Puppe, B., 1998. Effects of familiarity and relatedness on agonistic pair relationships in newly mixed domestic pigs. Applied Animal Behaviour Science 58, 233-239. SAS Institute Inc., 2005. SAS/STAT User's guide, Version 9.1. Schwarz, G., 1978. Estimating the dimension of a model. Annals of Statistics 6, 461-464. Spoolder, H.A.M., Geudeke, M.J., Van der Peet-Schwering, C.M.C., Soede, N.M., 2009. Group housing of sows in early pregnancy: A review of success and risk factors. Livestock Science 125, 1-14. Stookey, J.M., Gonyou, H.W., 1994. The effects of regrouping on behavioral and production parameters in finishing swine. Journal of Animal Science 72, 2804-2811.
58 Stukenborg, A., Traulsen, I., Puppe, B., Presuhn, U., Krieter, J., 2010a. Agonistic behaviour after mixing in pigs under commercial farm conditions. Applied Animal Behaviour Science (2010), doi:10.1016/j.applanim.2010.10.004. Stukenborg, A., Traulsen, I., Stamer, E., Puppe, B., Presuhn, U., Krieter, J., 2010b. Heritabilities of agonistic behavioural traits in pigs and their relationships within and between different age groups. submitted to Livestock Science. Tan, S.S.L., Shackleton, D.M., Beames, R.M., 1991. The effect of mixing unfamiliar individuals on the growth and production of finishing pigs. Animal Production 52, 201-206. Tuchscherer, M., Manteuffel, G., 2000. The effect of psycho stress on the immune system. Another reason for pursuing animal welfare (Review). Archiv für Tierzucht - Archives of Animal Breeding 43, 547-560. Tuchscherer, M., Puppe, B., Tuchscherer, A., Kanitz, E., 1998. Effects of social status after mixing on immune, metabolic, and endocrine responses in pigs. Physiology & Behavior 64, 353-360. Turner, S.P., White, I.M.S., Brotherstone, S., Farnworth, M.J., Knap, P.W., Penny, P., Mendl, M., Lawrence, A.B., 2006. Heritability of post-mixing aggressiveness in grower-stage pigs and its relationship with production traits. Animal Science 82, 615-620. Weng, R.C., Edwards, S.A., English, P.R., 1998. Behaviour, social interactions and lesion scores of group-housed sows in relation to floor space allowance. Applied Animal Behaviour Science 59, 307-316.
59 60 CHAPTER FOUR
The use of a lesion score as an indicator for agonistic behaviour in pigs
Andreas Stukenborg 1, Imke Traulsen 1; Eckhard Stamer 2, Birger Puppe 3, Ulrich Presuhn 4, Joachim Krieter 1
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, D-24098 Kiel, Germany
2TiDa Tier und Daten GmbH, D-24259 Westensee/Brux, Germany
3Leibniz Institute for Farm Animal Biology (FBN), D-18196 Dummerstorf, Germany
4farm concepts GmbH & Co. KG, D-23812 Wahlstedt, Germany
61 Abstract
The aim of the present study was to analyse whether a lesion score approach could be used as an indicator for agonistic behaviour in pigs. Behavioural patterns were observed directly after weaning (PIG-28; n = 827), 40 days later (growing pigs, PIG-68; n = 292) and immediately after mixing gilts (GILTS; n = 344), over a 48-hour period. For the front, middle and rear part of the body the extent of skin injuries was assessed with a lesion score (LS). The LS ranked from 0 (no wounds) to 4 (many, deep wounds) and was determined at the beginning (LSstart) and at the end (LSend) of the observation period. In all groups, the front body third was mostly stressed by fighting interactions. The highest correlations were found between the front LS and the overall fight time, with r = 0.50 (PIG-28), r = 0.52 (PIG-68) and r = 0.48 (GILTS). The LS for the front area did not increase for 26% (PIG-28), 39% (PIG-68) and 40% (GILTS) of the animals. According to this, the applied LS approach only reflected the agonistic behaviour of pigs to a certain extent.
Keywords: pig, agonistic behaviour, lesion score, correlation, age groups
62 1. Introduction
Agonistic behaviour in pigs especially occurs when unacquainted pigs are mixed with groups of acquainted ones. Fighting often results in the accumulation of skin lesions which can have detrimental effects on the welfare and longevity of the animals (Spoolder et al., 2000; Løvendahl et al., 2005). In recent studies, a skin lesion score (lesion score, LS) was applied as a proxy indicator for the aggressive potential of a pig (Turner et al., 2006). With regard to the results in Turner et al. (2008; 2009) the LS is a heritable trait and genetically correlated with aggressive behaviour. They concluded that a selection based on breeding values for a reduced LS is expected to reduce reciprocal aggression (Turner et al., 2008). However, most of the previous applied lesion scores have implemented detailed counting or assessment of each scratch (e.g. Erhardt et al., 1997; Weng et al., 1998; Spoolder et al., 2000; Turner et al., 2006). Thus, they seem to be infeasible to measure the LS under common breeding farm conditions. The aim of this study was to apply an easy and rapid lesion score application under commercial sow farm conditions. The behaviour and lesion score recordings were made in three age groups. Finally, the study examined whether the LS could be an indicator for agonistic behaviour in pigs. Therefore, correlations were estimated between the LS and agonistic behavioural traits.
2. Materials and Methods
2.1. Animals and Housing
The data collection was carried out on a closed herd sow farm of the German breeding company ‘Hülsenberger Zuchtschweine’ during October 2007 to April 2009. The herd composed of pure-bred Large-White (LW) sows and also crosses between Large-White and German Landrace (LW x DL). For the present study, female pigs from pure bred litters and crosses from LW x DL were considered. Behavioural observations from weaned (PIG-28) piglets and growing pigs (PIG-68) were made in the rearing pens. Every Wednesday piglets from one sow group with 36 litters on average were weaned (one-week rhythm) and moved to the rearing house where they were sorted by gender and body size. 29 piglets on average were housed in one rearing pen. The pens had dimensions of 5.10m x 2.10m. They had a fully slatted floor and were separated by 60cm high, closed partitions on the ground and two rods across. During the rearing period water was non-stop accessible on nipple drinkers. Feed was
63 provided ad libitum with a wet feeder. After the rearing period the pigs were taken to a fattening unit where the own performance test was implemented. The pigs were fed with a liquid feeding system three times a day. The third observation was carried out when the gilts (GILTS) were brought back to the sow farm directly after the first mixing in the pregnancy area. An average of 48 gilts were housed as a separate group in the gilt pen with dimensions of ca. 15.5m x 6.5m. The pen was divided into a feeding (ca. 43m 2) and a lying area (ca. 58m 2). The feeding area included the electronic sow feeding station and had a fully slatted floor. The lying area comprised a corridor with a fully slatted floor and six resting areas (each ca. 5.4m 2) with a solid floor. Water was accessible ad libitum on one nipple drinker and two water troughs.
2.2. Behavioural observation
Four rearing pens with female breeding piglets were filmed directly after weaning (PIG-28) using the HeiTel Player software program (HeiTel Digital Video GmbH, Kiel, Germany). The behavioural observations were made for 10 focus piglets per pen, which were randomly chosen and individually marked on their backs. The observation area excluded the area around the wet feeder and measured 3.20m x 2.10m. A second observation took place 40 days after weaning (growing pigs = PIG-68). A new group was created in equal shares from four pens which were previously housed at weaning. The same 40 focus pigs which had already been observed at weaning were re-marked, re-mixed and re-allocated to the four rearing pens. Thus, one out of four from the new group and also two or three from the marked pigs were already known to each other. The third observation was carried out in the gestation house (GILTS). Before the gilts were moved into the pen, each animal was individually marked. The recording period started immediately after housing the gilts. The behaviour of all animals inside the whole pen area was observed. The agonistic behaviour of each marked pig was evaluated with the aid of the videotapes. A single agonistic interaction was defined as a fight or a displacement with physical contact initiated by one individual and featuring aggressive behavioural elements, followed by any form of submissive behaviour performed by the opponent (Langbein and Puppe, 2004). The aggressor/receiver, the winner/loser of a fight and also the times (start time and finish time) were noted for each fight in which a marked pig was involved (for a more detailed description see Stukenborg et al. (2010a)). Based on the beginning and outcome of the fight, the values of ten agonistic behavioural traits were calculated for each observed pig (Table 1).
64 Table 1: Definitions of the agonistic behaviour traits Behavioural trait Unit Definition Fights no. of Number of all agonistic interactions per pig in which the pig was involved Time Fights seconds (s) The total time a marked pig was involved in agonistic interactions during the 48-hour observation period Aggressive Fights no. of The number of fights per pig in which the pig initiated the agonistic interaction Time Aggressive Fights seconds (s) The total time a marked pig was involved in self-initiated agonistic interactions during the 48-hour observation period Aggressions Received no. of The number of fights per pig in which the pig was subjected to (not the initiator of) the agonistic interaction Time Aggressions Received seconds (s) The total time a pig was subjected to agonistic interactions during the 48-hour observation period Fights Won no. of The number of fights per pig in which the pig won the agonistic interaction Time Fights Won seconds (s) The total time a marked pig was involved in won agonistic interactions during the 48-hour observation period Fights Lost no. of The number of fights per pig in which the pig lost the agonistic interaction Time Fights Lost seconds (s) The total time a pig was involved in lost agonistic interactions during the 48-hour observation period
2.3. Lesion score
A skin lesion score (LS) was assigned by one person to all pigs marked (LSstart) immediately before starting the 48-hour observation period. After 48 hours (PIG-68) or one week later (PIG-28 and GILTS), the LS was recorded again (LSend) for the observed pigs. The time differences between the first and second recordings (48 hours vs. one week) were necessary in order not to disturb the workflow on the farm. Three body regions were evaluated independently of each other: front (head, neck, shoulders and front legs), middle (flanks and back) and rear (rump, hind legs and tail) (modified according to Turner et al., (2006)). A detailed counting of each scratch on the pigs’ bodies was not carried out. Although such a procedure would probably have given an exact indication of the new wounds, the counting of
65 the lesions appeared to be too time-consuming for use under commercial farm conditions. The applied LS ranked from 0 to 4 as shown in Table 2.
Table 2: Description of the lesion score ranking Code Description
0 No skin injuries (no lesions on the whole body area)
1 Minor skin injuries (sporadically occurring) lesions on the body area) Middle skin injuries (several lesions allocated over the whole body area; no 2 widespread accumulation of injuries) Strong skin injuries (several lesions allocated over the whole body area with 3 accumulations of injuries at different areas of the skin 4 Very strong skin injuries (the whole body area is densely covered with lesions)
Finally, because of the missing of culled or diseased pigs for LSend recording and due to problems with the technical supply, both lesion scores as well as behavioural observations were ascertained for 827 (PIG-28), 292 (PIG-68) and 344 (GILTS) animals.
2.4 Statistical analysis
In order to reach approximately normal distributions, the behaviour data were transformed with a square root (for PIG-28) or log transformation Y = log e (1+observation) (for PIG-68 and Gilts). All further analyses were based on the transformed behaviour data. An analysis of fixed effects was performed with the MIXED procedure in SAS (2005). For model fitting, the information criteria of Hurvich and Tsai (1989) (AICC, Akaikes’s information criteria) and Schwarz (1978) (BIC, Bayesian information criteria) were considered. These values take into account the number of estimated parameters and favour less complex model variants. The model with the smallest AICC and BIC values was chosen without considering the underlying significance of fixed effects. For the behavioural traits, the final model for PIG-28 and PIG-68 included the fixed effect of the observation month and the random pen and litter effect. For GILTS behaviour, the fixed effect of the observation date was considered. For LSend, the same models were used for behavioural traits expanded by the fixed effect of LSstart. Pearson correlation coefficients were calculated between the residual of the behavioural traits and LSend (Proc CORR; SAS, 2005).
66 3. Results
Behavioural performance Most fighting interactions were observed for weaned piglets (Table 3). GILTS had the lowest median number of fights. However, the total fight time was considerably longer in comparison to growing pigs. Great differences in fighting behaviour were also seen within age groups. In all age classes at least one pig was not involved in agonistic interaction for the first 48 hours. In contrast, the most active animals had an overall fight time approximately 220 (PIG-28), 113 (PIG-68) and 108 (GILTS) minutes, respectively. A more detailed description of agonistic behaviour in weaned piglets and growing pigs is given in Stukenborg et al. (2010a).
Table 3: Median of the untransformed behavioural traits per animal for each age group. (Minimum = Min.; Maximum = Max.) PIG-28 PIG-68 GILTS
Median Min. / Max. Median Min. / Max. Median Min. / Max.
Fights 1) 39 0 / 188 8 0 / 86 6 0 / 33
Time Fights 2460 0 / 13171 99 0 / 6751 280 0 / 6481
Aggressive Fights 13 0 / 140 2 0 / 51 2 0 / 17 Time Aggressive 781 0 / 7083 18 0 / 2500 86 0 / 2992 Fights Aggressions Received 14 0 / 57 4 0 / 25 3 0 / 16 Time Aggressions 712 0 / 7076 47 0 / 3605 102 0 / 3640 Received Fights Won 5 0 / 78 0 0 / 41 1 0 / 16
Time Fights Won 214 0 / 2658 0 0 / 1344 33 0 / 2560
Fights Lost 5 0 / 63 2 0 / 15 2 0 / 13
Time Fights Lost 246 0 / 3039 22 0 / 1700 77 0 / 2190
1) Descriptions for the behaviour traits are given in Table 1.
Lesion score differences Figure 1 shows the distribution of the first and second LS for the body regions of each age group. In general an LSstart equal to zero was rarely observed for growing pigs and gilts. The LS for the front body region increased (LSstart < LSend) for 74% (PIG-28), 61% (PIG-68),
67 and 60% (GILTS) of the animals, respectively (results not presented). However, almost 50% of gilts had an LSstart greater than 2. Mostly small variations were observed between LSstart and LSend for the middle body area. The majority of the weaned piglets had a middle LSstart of 1. In comparison, PIG-68 and GILTS had higher starting values overall. About 50% (PIG- 28/PIG-68) and 56% (GILTS) of the pigs had no positive change in the middle LS. For the rear body area, most LS differences were observed for gilts. In contrast to the younger age groups, a considerable number of gilts had lesion scores of 3 or 4. However, an increased LS was only shown for nearly 29% (PIG-28), 33% (PIG-68) and 42% (GILTS) of the pigs, respectively.
70 front middle rear 60 PIG-28 50 LSstart 40 LSend 30
% piglets % 20 10 0 01234 01234 01234 60
50 PIG-68
40 30
% pigs % 20
10
0
60 01234 01234 01234 50 GILTS 40
30
% gilts % 20 10
0 01234 01234 01234 lesion score Figure 1: Distribution of the first and second LS for the body regions of each age group
Correlation between LS and behavioural parameters The highest correlation between agonistic behavioural parameters and the LSend in all age groups were shown for the front body area (Table 4). The maximum correlation coefficients were found with the overall fight time, r = 0.50 (PIG-28) (p < 0.05), r = 0.52 (PIG-68) (p < 0.05) and r = 0.48 (GILTS) (p < 0.05). Intermediate correlations were calculated for the
68 middle body region. The correlation coefficients ranked between r = -0.11 up to 0.26 (PIG- 28), r = 0.11 up to 0.29 (PIG-68) and r = 0.13 up to 0.30 (GILTS). No significant correlation for the rear area were found for weaned piglets. The highest significant relationships between the rear LS and behavioural parameters were calculated for GILTS (r = 0.33 for Time Fights and Time Fights Won (p < 0.05)).
Table 4: Pearson correlation coefficients between agonistic behavioural traits and the lesion score of three different body areas observed in three different age groups (weaned piglets = PIG-28, growing pig = PIG-68, gilts = GILTS) PIG-28 PIG-68 GILTS front middle rear front middle rear front middle rear
Fights -0.27* -0.14* -0.00 -0.49* -0.25* -0.17* -0.37* -0.21* -0.25*
Time Fights -0.50* -0.26* -0.01 -0.52* -0.27* -0.21* -0.48* -0.30* -0.33*
Aggressive Fights -0.16* -0.09* -0.01 -0.42* -0.28* -0.11* -0.33* -0.21* -0.27* Time Aggressive -0.34* -0.19* -0.01 -0.45* -0.29* -0.13* -0.37* -0.23* -0.30* Fights Aggressions -0.15* -0.07* -0.00 -0.41* -0.17* -0.15* -0.29* -0.15* -0.15* Received Time Aggressions -0.41* -0.20* -0.03 -0.46* -0.23* -0.20* -0.40* -0.24* -0.26* Received Fights Won -0.25* -0.17* -0.03 -0.48* -0.27* -0.17* -0.41* -0.23* -0.30*
Time Fights Won -0.33* -0.20* -0.00 -0.49* -0.27* -0.15* -0.45* -0.24* -0.33*
Fights Lost -0.16* -0.11* -0.02 -0.13* -0.11 * -0.11 * -0.15* -0.13* -0.04 *
Time Fights Lost -0.07* -0.02 * -0.01 -0.25* -0.19* -0.22* -0.37* -0.27* -0.22* *p < 0.05
4. Discussion
In general, the front body area was the most stressed body region of the pigs in all age groups. Furthermore, in all age classes the front third showed the highest correlations between LS and agonistic behaviour. These results are in accordance with McGlone (1985), who indicated that most of the bites were targeted mainly at the ears, face and neck. A comparison between the age groups showed that the correlation coefficients of growing pigs and gilts were rather similar in contrast to the findings of weaned piglets in which no significant relationship between the rear body area and agonistic behaviour was found. Differences between the age 69 groups might be caused through different fighting motivations and the resulting variations in the number of scratches. Previous results have indicated moderate correlation between the behaviour of growing pigs and gilts but only small relationships between the older age groups and weaned piglets (Stukenborg et al., 2010b). It is therefore assumed that the behaviour of weaned piglets is especially influenced by playful manners whereas the older age groups fight in order to establish a new dominance hierarchy. Regarding this, Stukenborg et al. (2010a) suggested that weaned piglets fought much more than growing pigs but their interactions were not so intensive, so a fight did not always end in injury. In all age groups, more than 50% of the pigs had no increased LS for the middle and rear body area. This fact and the consistently smaller relationships in comparison to the front LS suggest that an investigation of the rear and the middle lesion score seemed to be rather insignificant in relation to agonistic behaviour. Although the correlation between behavioural traits and the front LS were all significant and clearly showed that a higher involvement in agonistic interactions resulted in an increased LS, the correlation coefficients were at the maximum at the level of average. Thus, the applied LS approach did only reflect the agonistic behaviour of pigs to a certain extent. Furthermore, it must be considered that a lot of animals already had a high front LSstart. For these pigs the LS was limited concerning an indication of fighting behaviour since high fighting activity could not lead to a high difference between LSstart and LSend, Concerning gilts, even more than 20% had a front LSstart equal to 4. For these animals further scratches in the observation period due to fighting interactions could not result in a higher LS. Thus, it must be considered that before keeping the gilts in the pregnancy area, the animals were taken from the fattening unit to the sow farm and were subsequently washed as a whole group. During this time the animals could already have been involved in agonistic interactions but the added wounds were assigned to the LSstart. For this reason, in further investigations the LSstart should be noted before the animals leave their familiar group and start fighting with unacquainted pigs. Previously applied LS approaches have avoided this problem by counting the number of lesions at the beginning and at the end of the observation period (Weng et al., 1998; Spoolder et al., 2000; Turner et al., 2006) or they have tried to count only fresh lesions which were received during the observation time (Erhard et al., 1997; Turner et al., 2000; Turner et al., 2009). Turner et al. (2009) found genetic correlations between the anterior LS and the duration in reciprocal aggression of r = 0.67. In comparison to these findings which include laborious data collections, the present results, with correlations of approximately r = 0.50 in all age groups, are quite promising when considering the short and easy lesion score ascertainment.
70 5. Conclusion
In contrast to the middle and rear body region, the front body area is highly stressed by agonistic interactions in all age groups. The estimated moderate correlation coefficient between the applied front LS and the observed agonistic behavioural traits indicated that a quick and basic approach can be used for a lesion score ascertainment. However, in further studies the LSstart should be noted before the animals leave their familiar pen in order to include all newly received wounds.
References
Erhard, H.W., Mendl, M., Ashley, D.D., 1997. Individual aggressiveness of pigs can be measured and used to reduce aggression after mixing. Applied Animal Behaviour Science 54, 137-151. Hurvich, C.M., Tsai, C.L., 1989. Regression and Time-Series Model Selection in Small Samples. Biometrika 76, 297-307. Langbein, J., Puppe, B., 2004. Analysing dominance relationships by sociometric methods - a plea for a more standardised and precise approach in farm animals. Applied Animal Behaviour Science 87, 293-315. Løvendahl, P., Damgaard, L.H., Nielsen, B.L., Thodberg, K., Su, G., Rydhmer, L., 2005. Aggressive behaviour of sows at mixing and maternal behaviour are heritable and genetically correlated traits. Livestock Production Science 93, 73-85. McGlone, J.J., 1985. A quantitative ethogram of aggressive and submissive behaviors in recently regrouped pigs. Journal of Animal Science 61, 559-565. SAS Institute Inc., 2005. SAS/STAT User's guide, Version 9.1. Schwarz, G., 1978. Estimating the dimension of a model. Annals of Statistics 6, 461-464. Spoolder, H.A.M., Edwards, S.A., Corning, S., 2000. Aggression among finishing pigs following mixing in kennelled and unkennelled accommodation. Livestock Production Science 63, 121-129. Stukenborg, A., Traulsen, I., Puppe, B., Presuhn, U., Krieter, J., 2010a. Agonistic behaviour after mixing in pigs under commercial farm conditions. Applied Animal Behaviour Science (2010), doi:10.1016/j.applanim.2010.10.004.
71 Stukenborg, A., Traulsen, I., Stamer, E., Puppe, B., Presuhn, U., Krieter, J., 2010b. Heritabilities of agonistic behavioural traits in pigs and their relationships within and between different age groups. submitted to Livestock Science. Turner, S.P., Ewen, M., Rooke, J.A., Edwards, S.A., 2000. The effect of space allowance on performance, aggression and immune competence of growing pigs housed on straw deep-litter at different group sizes. Livestock Production Science 66, 47-55. Turner, S.P., Farnworth, M.J., White, I.M.S., Brotherstone, S., Mendl, M., Knap, P., Penny, P., Lawrence, A.B., 2006. The accumulation of skin lesions and their use as a predictor of individual aggressiveness in pigs. Applied Animal Behaviour Science 96, 245-259. Turner, S.P., Roehe, R., D'Eath, R.B., Ison, S.H., Farish, M., Jack, M.C., Lundeheim, N., Rydhmer, L., Lawrence, A.B., 2009. Genetic validation of postmixing skin injuries in pigs as an indicator of aggressiveness and the relationship with injuries under more stable social conditions. Journal of Animal Science 87, 3076-3082. Turner, S.P., Roehe, R., Mekkawy, W., Farnworth, M.J., Knap, P.W., Lawrence, A.B., 2008. Bayesian analysis of genetic associations of skin lesions and behavioural traits to identify genetic components of individual aggressiveness in pigs. Behavior Genetics 38, 67-75. Weng, R.C., Edwards, S.A., English, P.R., 1998. Behaviour, social interactions and lesion scores of group-housed sows in relation to floor space allowance. Applied Animal Behaviour Science 59, 307-316.
72 GENERAL DISCUSSION
The aim of the present study was to analyse the agonistic behaviour of pigs in different age groups under common farm conditions. A description was given concerning the time course as well as the number and duration of agonistic interactions. In addition to the heritability estimates for agonistic behavioural traits, the study analysed the relationship between aggressiveness and growth as well as between aggressiveness and reproductive performance traits. Finally, a lesion score approach was investigated to determine whether it could be used as a proxy indicator for the aggressiveness potential of pigs.
Data collection under commercial farm conditions
In general, the mixing of pigs on a sow farm occurs directly after weaning (weaned piglets), when pigs leave the rearing pens and move to the finishing unit (growing pigs) and finally when sows are housed in the pregnancy area. In order to illustrate the ontogenesis of agonistic behaviour over three age classes, the data collection was performed on a commercial farm where all three regrouping times could be recorded. However, several difficulties were associated with the data collection under practical farm conditions. For example, due to the large pens and group sizes the area around the wet feeder was not observed and an observation of all 29 pigs in the rearing pens was not possible. Only ten pigs out of every rearing pen were marked in order to follow of their fighting behaviour. Hence, an exact analysis of the whole dominance hierarchy in the pen according to Langbein and Puppe (2004) was not feasible. Because of the commonly applied regrouping practice, pigs also known to each other were mixed together in one pen. This probably influenced the overall fighting level in the pens because acquainted pigs seem to have shorter fighting times among themselves (Jensen and Yngvesson, 1998). Furthermore, the farm circumstances did not allow the implementation of a special behavioural test. A resident-intruder test for example, which shows the level of aggression between a resident pig which remains in the home pen and an introduced intruder (Erhard and Mendl, 1997), would have been too time-consuming and did not fit in the workflow. This also applies to animal weighting. Various investigations have indicated the influence of body size on the agonistic behaviour in pigs (Olesen et al., 1996; D'Eath, 2002; Turner et al., 2006). Therefore, it would have been beneficial to record the body weight of each pig. However, this was not possible due to the fixed work schedule on the farm.
73 Despite these difficulties, the present findings indicate a circadian rhythm in agonistic behaviour and also that fighting behaviour persisted over the whole 48-hour observation period after mixing. Furthermore, considerable differences in the agonistic behaviour of pigs were found within and between the age groups. Considering these results, the present study clearly indicated that video equipment can be used to analyse the phenotypes of agonistic behaviour in pigs under commercial farm conditions. Generally, data recording under common farm conditions seems to be helpful with regard to a possible implementation of agonistic behavioural traits in pig breeding because no genotype environment interactions are to be expected. In contrast, such interactions normally occur during special test situations where the test environment does not reflect the common environment.
Indicators for agonistic behaviour
In the literature, various tests have already been conducted revealing the behaviour of pigs. A human approach test for example was applied by several authors (Thodberg et al., 1999; Janczak et al., 2003) in order to show the animals’ fear of humans. During these tests the animals were kept alone and their reactions were observed after a human entered the test pen. Brown et al. (2009) described a novel object test which provided a measure of fear and exploratory behaviour directed towards a novel object in the pen. They recorded the latency for the contact with the object for each pig. Another commonly applied test for piglets is the backtest. For this trial, the piglets are restrained on their backs. Classification of the pigs is based on the number of escape attempts made during 60 s (Bolhuis et al., 2005). Although these tests do not directly indicate the aggressive potential of a pig, they reflect an indication of temperament in the animals. It could be conceivable that pigs which show fear of humans or novel objects are also those which would avoid fighting interactions and would show more submissive behaviour. In relation to this, Brown et al. (2009) indicated that pigs which were faster to approach a human tended to be more aggressive at mixing. The results from Bolhuis et al. (2005) showed that animals which were highly resistant during the backtest were also significantly more aggressive after regrouping. With regard to these findings, an application of a particular temperament test could also reflect the agonistic behaviour of pigs to a certain extent. However, trials in which aggressiveness is directly observed would probably give a more precise description of the agonistic pig behaviour. In a resident-intruder test, Erhard and Mendl (1997) for example analysed the aggressive behaviour between a resident pig which remained in the home pen and an intruder pig. Løvendahl et al. (2005) tried to provoke
74 fighting behaviour during the 30-minute observation period by keeping 5 to 10 sows in a testing arena. Such an induced stress situation lays the focus especially on the agonistic behaviour. Due to the relatively small number of pigs and the possibility to define environmental effects, such as space allowance for example, the agonistic behaviour of each pig can clearly be analysed. Based on the test situation of Løvendahl et al. (2005), intermediate heritabilities were estimated for performed aggression traits of sows. Conversely, in the present study the heritability for the corresponding traits concerning the initiation of fights were lower than h 2 = 0.03. As discussed in Chapter Two , the considerable differences between the heritability estimates might be caused by differences in the behavioural ascertainment. In the present study, the gilts had a high space allowance and visual barriers in the pen. Thus, a clear estimation of the sows’ agonistic potential was probably not always stated because the gilts could completely avoid agonistic interactions. By contrast, a temperament test which induces fighting activity can probably better reflect the agonistic potential of gilts. However, it should be considered that most of the applied test circumstances will never occur on commercial pig farms and nobody has yet analysed whether the agonistic behaviour of a sow during a test situation is completely comparable with the agonistic behaviour shown in a common pregnancy area. Therefore, the benefit of such temperament tests for pig breeding and welfare seems to be questionable. In addition to this, a general problem of the tests mentioned is the relatively high amount of time necessary caused by the direct behavioural observations over a definite period of time. The use of a lesion score approach to indicate agonistic behaviour can probably reduce the need for time-consuming data collection. In the present study, an easy and short lesion score assessment was applied by scoring the number of scratches on the skin on a scale from 0 to 4. Despite this relatively subjective grading, the phenotypic correlations between lesion score and agonistic behaviour, estimated in Chapter Four , were on the same level with Turner et al. (2009), who carried out a more exact but also more laborious lesion scoring. Regarding the advantages and disadvantages of the different behavioural tests and lesion scores, the applied lesion score in the present study seems to be one of the simplest approaches to indicate the aggressiveness of pigs. Nevertheless, it must be considered that the present lesion score as well as the other temperament tests only reflected the agonistic behaviour of pigs to a certain extent.
75 Ontogenesis of agonistic behaviour
An important aspect of the present study was to show the ontogenesis of agonistic pig behaviour. Therefore the agonistic behaviour was analysed directly after mixing in three age groups, which were chosen in accordance with the common regrouping times in pig husbandry. In order to bring about a reduction in agonistic behaviour as a breeding goal in pig breeding and to find suitable recording times for agonistic behaviour it is important to know whether a young piglet with a high aggressive potential is also highly aggressive in the older age groups. The results in Chapter Two indicate no clear relationship between the behaviour of weaned piglets and the older age groups since the agonistic behaviour of the piglets is also influenced by playful manners. Thus, it would make no sense to observe their behaviour directly after weaning in order to have calmer fattening pigs or sows. The preferred observation time for consideration in pig breeding seems to be after the rearing period because of the moderate heritabilities for the agonistic behavioural traits at this point of time. Furthermore, concerning the positive relationships between growing pigs and gilts, the use of the behavioural data after 68 days could probably decrease the aggressiveness in the fattening units as well as the aggressiveness of sows.
Changing agonistic behaviour - welfare and ethical aspects
The negative impact of aggressive behaviour in pigs on welfare and production traits has been described in different studies: it has been suggested that breeding against agonistic behaviour could probably help to increase the welfare of pigs (Løvendahl et al., 2005; D'Eath et al., 2009; Turner et al., 2010). Certainly, a decrease in agonistic behaviour after mixing would lead to fewer skin lesions and a lower risk for immunosuppressive effects. This in turn would have positive economic effects for pig husbandry. Nevertheless, while breeding for behaviour is potentially beneficial, ethical concerns have been raised because an adaption to the environment can probably lead to a loss of the pigs’ ‘naturalness’ and their integrity (D'Eath et al., 2010). According to D’Eath et al. (2010), selection against certain behaviours could carry a risk of breeding animals which are generally non-reactive. Regarding this, it has been shown pigs which are less aggressive at mixing are also less aggressive during the weighing procedure (D'Eath et al., 2009). Thus, through genetic correlations, various behaviours could be changed even if only the aggressiveness is to be decreased. In pig husbandry, farmers want to have calmer animals in their pens which are rather less involved in agonistic interactions. However, fighting interactions can usually be seen in the wild (Graves, 1984) and so the
76 fighting behaviour seems to be one special indication of the natural behaviour of domestic pigs. Therefore, a decrease in agonistic behaviour would also decrease the ‘naturalness’ of the pigs. Referring to Fraser (1999), the possibility to reveal normal behaviour is in turn an important aspect for the welfare of the animals. In course of the more and more increasing public interest in animal welfare, it is important to consider such aspects in pig breeding in order not to decrease the public reputation of the overall pig husbandry. Nevertheless, due to the possibility to reduce the number of wounds, breeding against agonistic behaviour seems to be also a way of breeding for improved welfare and according to Kanis et al. (2004) breeding for better animal welfare seems to be ethically justifiable. The present study considered an economic aspect for breeding against aggressive behaviour by analysing the relationships between agonistic behaviour and common performance traits in pig husbandry ( Chapter Three ). No correlations were found which indicated an unfavourable interaction between agonistic behaviour and growth and reproductive performance traits, respectively. However, concerning breeding, further studies are necessary to also estimate genetic correlations. With regard to ethical reasons, it must be observed whether a reduction in agonistic behaviour leads to conspicuous changes in the overall behaviour and therefore probably to a negative effect in pig welfare.
Recommendations
Based on the results of the present study, certain suggestions can be made which may probably help to improve further investigations dealing with agonistic behaviour in pigs. In general, the agonistic interactions between weaned piglets can be ignored in further studies. Their agonistic behaviour seemed to be highly influenced by playful manners and was not correlated with the behavioural traits of the older age groups. In contrast, the agonistic behaviour of growing pigs and sows is becoming more important due to legal regulations. The group housing of pregnant sows for example will be required by law as of 2013. Currently, in order to avoid the castration of piglets, there is increasing interest in the fattening of boars, which in turn might be associated with an increased fighting level in the pens. In order to obtain information more related to practice, further studies should be made under common farm conditions. With relation to the present findings a video observation of the agonistic behaviour might be required only in essential observation periods; e.g. in the daytime when most of the fights occur (Chapter One ). However, it should be analysed whether an overall observation period of 72 or 96 hours for example could give more exact behavioural
77 descriptions than a 48-hour observation period as applied in the present study. In addition to a lesion score, further practical indicators should be examined concerning their expressiveness of agonistic behaviour. Corresponding to the study of D’Eath et al. (2009), behaviour could be observed during a routine weighing procedure or when sows are washed before they are kept in the farrowing area. A video observation had to be carried out simultaneously in the study in order to compare the exact behavioural valuations with the results of the proxy indicators. Due to the lack of investigations available in the literature, it is still uncertain as to whether a greater consideration of agonistic behaviour in pig breeding would have negative effects on the common performance traits. Therefore, additional studies are necessary in order to estimate genetic correlations between agonistic behaviour and growth and reproductive performance, respectively. Also the relationships with maternal behaviour patterns or behaviour towards humans should be analysed. In addition, investigations into whether management provisions or housing conditions affect the occurrence of agonistic behaviour should be carried out. In general, a reduction of agonistic behaviour could be obtained by increasing group size and pen space, respectively. Schmolke et al. (2004) suggested that due to the large group size the pigs can avoid aggression or choose with whom they fight. Spoolder et al. (2009) indicated that in small sow groups with a low space allowance sows cannot express submissive behaviour and are more stressed compared to sows in large groups with the same space allowance per sow. A structured pen design for pregnant sows could also help to avoid agonistic interactions. Providing visual barriers for example allows the animals to hide and flee from larger, feared sows (Arey and Edwards, 1998; Weng et al., 1998; Spoolder et al., 2009). Furthermore, it seems to be helpful to provide bedding and straw during mixing times in order to distract the animals. Indeed, aggression would not be reduced by bedding but the risk of associated leg problems would perhaps decrease (Spoolder et al., 2009). Additionally, Spoolder et al (2009) suggested that the provision of straw promotes the exploratory behaviour of the animals and reduces the development of stereotypic manners.
78 References
Arey, D.S., Edwards, S.A., 1998. Factors influencing aggression between sows after mixing and the consequences for welfare and production. Livestock Production Science 56, 61-70. Bolhuis, J.E., Schouten, W.G.P., Schrama, J.W., Wiegant, V.M., 2005. Individual coping characteristics, aggressiveness and fighting strategies in pigs. Animal Behaviour 69, 1085-1091. Brown, J.A., Dewey, C., Delange, C.F.M., Mandell, I.B., Purslow, P.P., Robinson, J.A., Squires, E.J., Widowski, T.M., 2009. Reliability of temperament tests on finishing pigs in group-housing and comparison to social tests. Applied Animal Behaviour Science 118, 28-35. D'Eath, R.B., 2002. Individual aggressiveness measured in a resident-intruder test predicts the persistence of aggressive behaviour and weight gain of young pigs after mixing. Applied Animal Behaviour Science 77, 267-283. D'Eath, R.B., Conington, J., Lawrence, A.B., Olsson, I.A.S., Sandøe, P., 2010. Breeding for behavioural change in farm animals: practical, economic and ethical considerations. Animal Welfare 19, 17-27. D'Eath, R.B., Roehe, R., Turner, S.P., Ison, S.H., Farish, M., Jack, M.C., Lawrence, A.B., 2009. Genetics of animal temperament: aggressive behaviour at mixing is genetically associated with the response to handling in pigs. Animal 3, 1544-1554. Erhard, H.W., Mendl, M., 1997. Measuring aggressiveness in growing pigs in a resident- intruder situation. Applied Animal Behaviour Science 54, 123-136. Fraser, D., 1999. Animal ethics and animal welfare science: bridging the two cultures. Applied Animal Behaviour Science 65, 171-189. Graves, H.B., 1984. Behavior and ecology of wild and feral swine (sus scrofa). J. Anim Sci. 58, 482-492. Janczak, A.M., Pedersen, L.J., Bakken, M., 2003. Aggression, fearfulness and coping styles in female pigs. Applied Animal Behaviour Science 81, 13-28. Jensen, P., Yngvesson, J., 1998. Aggression between unacquainted pigs - sequential assessment and effects of familiarity and weight. Applied Animal Behaviour Science 58, 49-61.
79 Kanis, E., van den Belt, H., Groen, A.F., Schakel, J., de Greef, K.H., 2004. Breeding for improved welfare in pigs: a conceptual framework and its use in practice. Animal Science 78, 315-329. Langbein, J., Puppe, B., 2004. Analysing dominance relationships by sociometric methods - a plea for a more standardised and precise approach in farm animals. Applied Animal Behaviour Science 87, 293-315. Løvendahl, P., Damgaard, L.H., Nielsen, B.L., Thodberg, K., Su, G., Rydhmer, L., 2005. Aggressive behaviour of sows at mixing and maternal behaviour are heritable and genetically correlated traits. Livestock Production Science 93, 73-85. Olesen, L.S., Nygaard, C.M., Friend, T.H., Bushong, D., Knabe, D.A., Vestergaard, K.S., Vaughan, R.K., 1996. Effect of partitioning pens on aggressive behavior of pigs regrouped at weaning. Applied Animal Behaviour Science 46, 167-174. Schmolke, S.A., Li, Y.Z.Z., Gonyou, H.W., 2004. Effects of group size on social behavior following regrouping of growing-finishing pigs. Applied Animal Behaviour Science 88, 27-38. Spoolder, H.A.M., Geudeke, M.J., Van der Peet-Schwering, C.M.C., Soede, N.M., 2009. Group housing of sows in early pregnancy: A review of success and risk factors. Livestock Science 125, 1-14. Thodberg, K., Jensen, K.H., Herskin, M.S., 1999. A general reaction pattern across situations in prepubertal gilts. Applied Animal Behaviour Science 63, 103-119. Turner, S.P., D'Eath, R.B., Roehe, R., Lawrence, A.B., 2010. Selection against aggressiveness in pigs at re-grouping: practical application and implications for long-term behavioural patterns. Animal Welfare 19, 123-132. Turner, S.P., Roehe, R., D'Eath, R.B., Ison, S.H., Farish, M., Jack, M.C., Lundeheim, N., Rydhmer, L., Lawrence, A.B., 2009. Genetic validation of postmixing skin injuries in pigs as an indicator of aggressiveness and the relationship with injuries under more stable social conditions. Journal of Animal Science 87, 3076-3082. Turner, S.P., White, I.M.S., Brotherstone, S., Farnworth, M.J., Knap, P.W., Penny, P., Mendl, M., Lawrence, A.B., 2006. Heritability of post-mixing aggressiveness in grower-stage pigs and its relationship with production traits. Animal Science 82, 615-620. Weng, R.C., Edwards, S.A., English, P.R., 1998. Behaviour, social interactions and lesion scores of group-housed sows in relation to floor space allowance. Applied Animal Behaviour Science 59, 307-316.
80 GENERAL SUMMARY
This thesis focuses on the analysis of agonistic behaviour in pigs under commercial farm conditions. Agonistic interactions were examined in three age groups in order to give hints concerning the ontogenesis of agonistic behaviour. In addition, a lesion score was used to analyse its possible application as a proxy indicator of agonistic behaviour. Data recording was performed on a closed herd sow farm from October 2007 to April 2009. The behavioural observations were made over a 48-hour period during three common mixing times in pig husbandry: directly after weaning (weaned piglets), 40 days later (growing pigs) and immediately after mixing gilts (gilts). Only female pigs from pure-bred litters (Large- White) and crosses from Large-White x German Landrace were observed. Agonistic interactions were evaluated by noting the times (start and finish), the aggressor/receiver and the winner/loser of a fight. Based on these data different traits were calculated which described the aggressive or submissive behaviour of the pigs. Measurements for growth performance traits (live daily gain and backfat depth) were taken seven weeks before the gilts were mixed in the pregnancy area. In order to consider reproductive performance traits, the number of total born piglets, the number of piglets born alive and the number of stillborn piglets was ascertained for each litter.
In Chapter One , the fighting behaviour of weaned piglets and growing pigs was described for the time after mixing. Furthermore, a dominance index and a lesion score were presented in order to show their expressiveness with regard to agonistic behaviour. Weaned piglets showed an overall higher level of agonistic interactions than growing pigs but a circadian rhythm was shown for both age groups. Significantly more fights were observed during the daytime than during the night. The overall fighting level persisted over the whole 48-hour observation. With the aid of the dominance index, the pigs were divided in dominant animals, showing more wins than defeats and subordinate animals which lost more fights than won them. The dominant pigs were engaged in significantly more agonistic interactions, had a longer total fight time and initiated more fights than subordinate pigs. Regarding the lesion score, it became apparent that with an increasing involvement in fights the number of lesions increased. The dominance index as well as the lesion score seemed to be feasible methods to analyse fighting behaviour under commercial farm conditions.
81 The aim of Chapter Two was to estimate the heritabilities for different agonistic behaviour traits and to compare the different agonistic behaviour between the age groups. The highest heritability for weaned piglets was estimated for the duration spent in initiated fights (h 2 = 0.20). For growing pigs, intermediate heritabilities were estimated, with the highest heritability for the number of fights won (h 2 = 0.37). Concerning gilts, substantial heritabilities were only found for the number of aggressions received (h 2 = 0.42), the time spent in aggressions received (h 2 = 0.15), the number of fights lost (h 2 = 0.13) and the time spent in fights lost (h 2 = 0.09). Only small correlations coefficients were estimated between the agonistic behavioural traits of weaned piglets and the older age groups, while moderate correlations were found between the behavioural traits of growing pigs and gilts. Thus, the results indicate that the more aggressive growing pigs tended to be the more aggressive sows.
The objective of Chapter Three was to analyse the relationship between agonistic pig behaviour and the common performance traits (growth and reproduction performance) in pig breeding. No clear relationship was found between the agonistic behaviour of weaned piglets and their subsequent growth performance. For growing pigs, the results indicated a positive relationship between backfat depth and the agonistic behaviour. Concerning gilts, animals with a higher live daily gain seemed to be more involved in agonistic interactions. No significant correlations were estimated between the agonistic behaviour of newly mixed gilts and the number of total born piglets, the number of piglets born alive as well as the number of stillborn piglets. Therefore, the fighting level of the animals had no influence on their subsequent fertility performance.
Chapter Four presents the results of examinations into whether a lesion score approach could be used as an indicator for agonistic behaviour in pigs. The lesion score was ascertained for the front, middle and rear body area and ranked from 0 (no wounds) to 4 (many, deep wounds). The scoring of the animals was carried out twice, at the beginning and at the end of the observation period. The results clearly indicate that in all age groups the front body region was mostly stressed by the pigs’ involvement in agonistic interactions. However, during the two evaluation times the lesion score of the front area did not increase for 26% (weaned piglets), 39% (growing pigs) and 40% (gilts) of the animals. The highest correlations were found between the front LS and the overall fight time, with r = 0.50 (weaned piglets), r = 0.52 (growing pigs) and r = 0.48 (gilts). To a certain extent these findings indicate that a quick and easy lesion score approach can be used to reflect the agonistic behaviour of pigs.
82 ZUSAMMENFASSUNG
Das Ziel der vorliegenden Arbeit bestand darin das agonistische Verhalten von Schweinen unter praktischen Betriebsbedingungen zu analysieren. Um Aussagen bezüglich der Ontogenese des agonistischen Verhaltens treffen zu können wurde dieses in drei unterschiedlichen Altersklassen erfasst. Darüber hinaus wurde untersucht inwieweit die Bonitur von Hautverletzungen (Lesion Score) Rückschlüsse auf das agonistische Verhalten der Tiere zulässt. Die Datenerhebung fand von Oktober 2007 bis April 2009 auf einem Schweinezuchtbetrieb statt. Das Verhalten der Tiere wurde per Videoaufzeichnung über 48 Stunden während der drei für die Schweinehaltung typischen Umgruppierungsphasen beobachtet: direkt nach dem Absetzen, 40 Tage später vor der Einstallung in die Mast und unmittelbar nach dem Zusammenstallen der Jungsauen. Nur weibliche Tiere der Rasse Large-White bzw. Kreuzungstiere (Large-White x Deutsche Landrasse) wurden berücksichtigt. Bei der Auswertung der Videoaufzeichnungen wurden Start- und Endzeit der einzelnen Kämpfe erfasst. Ferner wurden der Gewinner und Verlierer notiert sowie welches Tier den Kampf initiiert hatte bzw. angegriffen worden war. Anhand dieser Daten konnten verschiedene Verhaltensmerkmale ermittelt werden, die das aggressive bzw. unterwürfige Verhalten der Tiere beschreiben. Sieben Wochen vor dem Einstallen der Jungsauen in den Wartestall wurden die Merkmale der Eigenleistungsprüfung (Lebenstagszunahme und Rückenspeckdicke) erhoben. Zur Berücksichtigung der späteren Fruchtbarkeitsleistungen der Sauen standen für jeden Wurf die Informationen über die Anzahl lebend geborener Ferkel, die Anzahl gesamt geborener Ferkel sowie die Anzahl der totgeborenen Ferkel zur Verfügung.
Im ersten Kapitel wurde das agonistische Verhalten von abgesetzten Ferkeln und Masttieren beschrieben. Insgesamt waren die abgesetzten Ferkel deutlich mehr in agonistische Interaktionen involviert als die Masttiere. Ein Tag-Nacht-Rhythmus konnte aber für beide Altersgruppen nachgewiesen werden. Zum einen zeigte sich, dass tagsüber wesentlich mehr gekämpft wurde als nachts und zum anderen wurde deutlich, dass, anders als in vorherigen Studien beschrieben, die Anzahl der Kämpfe am zweiten Tag nach der Umgruppierung nicht zurückging. Mit Hilfe eines Dominanzindexes wurden die Tiere in dominante (mehr Siege als Niederlagen) und subdominante Tiere (mehr Niederlagen als Siege), eingeteilt. Die dominaten Tiere initiierten mehr Kämpfe und waren häufiger wie auch insgesamt länger an agonistischen Interaktionen beteiligt. Der Lesion Score zeigte, dass ein Anstieg der Hautverletzungen mit
83 einer vermehrten Kampfbeteiligung einherging. Die Ergebnisse weisen darauf hin, dass sowohl der Dominanzindex als auch ein Lesion Score unter praktischen Bedingungen grundsätzlich zur Analyse des agonistischen Verhaltens genutzt werden kann.
Ziel des zweiten Kapitels war es die Heritabilitäten der agonistischen Verhaltensmerkmale zu berechnen. Darüber hinaus wurde untersucht inwieweit das Verhalten einer Altersgruppe mit anderen Altersklassen korrelierte. Für das Merkmal ‚Kampfzeit in initiierten Kämpfen pro Tier’ konnte bei den abgesetzten Ferkeln die höchste Heritabilität (h 2 = 0,20) festgestellt werden. Für die Masttiere wurden mittlere Heritabilitäten geschätzt. Hier konnte für das Merkmal ‚Anzahl gewonnener Kämpfe pro Tier’ mit h 2 = 0,37 die höchste Heritabilität ermittelt werden. Bei den Jungsauen hatten die Merkmale ‚Anzahl unterworfener Kämpfe pro Tier’, ‚Kampfzeit in unterworfenen Kämpfen pro Tier’, ‚Anzahl verlorene Kämpfe pro Tier’ und ‚Kampfzeit in verlorenen Kämpfen pro Tier’ Heritabilitäten von h 2 = 0,42; 0,15; 0,13 und 0,09. Für alle anderen Merkmale lagen die Schätzwerte bei h 2 ≤ 0,03. Des Weiteren zeigte sich, dass das agonistische Verhalten der abgesetzten Ferkel nur gering mit dem Verhalten der älteren Gruppen korrelierte. Im Gegensatz dazu konnten mittlere Korrelationen zwischen den Verhaltensmerkmalen der Masttiere und Sauen festgestellt werden. Die Ergebnisse deuten darauf hin, dass Tiere, die nach der Aufzuchtphase ein erheblich aggressives Verhalten zeigen auch als Sau ein hohes aggressives Potential aufweisen.
Ziel des dritten Kapitels war es die Beziehungen zwischen dem agonistischen Verhalten und den traditionellen Selektionskriterien in der Schweinezucht zu untersuchen. Zwischen dem agonistischen Verhalten von abgesetzten Ferkeln konnte kein eindeutiger Zusammenhang zu den Merkmalen der Eigenleistungsprüfung festgestellt werden. Bei den Masttieren war die Rückenspeckdicke positiv mit dem agonistischen Verhalten korreliert. Im Bezug auf die Jungsauen zeigen die Ergebnisse, dass Tiere mit einer höheren Lebenstagszunahme verstärkt an agonistischen Interaktionen beteiligt waren. Das agonistische Verhalten der Jungsauen korrelierte jedoch mit keinem der drei betrachteten Fruchtbarkeitsmerkmale. Somit scheint die Beteiligung an agonistischen Interaktionen keinen Einfluss auf die spätere Fruchtbarkeitsleistung zu haben.
In Kapitel 4 wurde untersucht inwieweit die Bonitur der Hautverletzungen (Lesion Score) zur Beschreibung des agonistischen Verhaltens der Tiere herangezogen werden kann. Der Lesion Score wurde für die vordere, mittlere und hintere Körperregion der Tiere ermittelt und
84 rangierte von 0 (keine Wunden) bis 4 (sehr viele, tiefe Wunden). Die Bewertung erfolgte zu Beginn und am Ende einer jeden Beobachtungsperiode. Die Ergebnisse zeigten deutlich, dass sich die Tiere beim Kämpfen insbesondere in der vorderen Körperpartie verletzten. Allerdings wurde bei 26% der abgesetzten Ferkel, 39% der Masttiere und 40% der Jungsauen kein Anstieg des Lesion Scores zwischen dem ersten und zweiten Erfassungszeitpunkt festgestellt. Von den Verhaltensmerkmalen war die ‚gesamte Kampfzeit pro Tier’ am höchsten mit dem Lesion Score korreliert. Hier lagen die Korrelationen bei r = 0,50 für abgesetzte Ferkel, r = 0,52 bei der Einstallung in die Mast und r = 0,48 für Jungsauen. Die Ergebnisse zeigen, dass durch eine einfache und schnelle Bonitur der Hautverletzungen (Lesion Score) Rückschlüsse auf das agonistische Verhalten der Tiere gezogen werden können.
85 86 DANKSAGUNG
An dieser Stelle möchte ich mich bei den Personen bedanken, die zum Gelingen dieser Arbeit beigetragen haben.
Mein besonderer Dank gilt Herrn Prof. Dr. J. Krieter für die Überlassung des interessanten Themas, die wissenschaftliche Betreuung, die gewährten Freiräume während der Erstellung der Arbeit sowie für die Möglichkeit meine Ergebnisse auf Tagungen im In- und Ausland vorstellen zu dürfen.
Für die Übernahme des Koreferats bedanke ich mich bei Herrn Prof. Dr. G. Thaller.
Ausdrücklich möchte ich mich bei Familie Wilkens aus Düdenbüttel sowie Herrn Dr. U. Presuhn (farm concepts) für die Möglichkeit der Datenerhebung bzw. der Datenbereitstellung bedanken.
Bei Herrn Dr. B. Puppe vom Leibniz-Institut für Nutztierbiologie (FBN) aus Dummerstorf bedanke ich mich für die fachliche Unterstützung im verhaltensphysiologischen Bereich.
Ein großes Dankeschön geht an Frau Dr. I. Traulsen und Herrn Dr. E. Stamer für die vielen wertvollen Diskussionen und der unermüdlichen Hilfe bei den statistischen Auswertungen.
Für die finanzielle Unterstützung danke ich der H. Willhelm Schaumann Stiftung.
Bei allen Kollegen und Kolleginnen möchte ich mich für die schöne Zeit am Institut bedanken. Vor allem danke ich allen jetzigen und ehemaligen „Containermitbewohnern“ für das angenehme Arbeitsklima und den Spaß den wir während der - mal mehr, mal weniger - fachlichen Diskussionen hatten. Stephan danke ich für die tolle Zeit im gemeinsamen Büro und der daraus entstandenen Freundschaft. Dies gilt auch für Andi, Thomas, Jan und Ewald. Unsere Abende im ‚Trotzdem‘ und die damit einhergehenden Fußballfachgespräche waren immer eine sehr willkommen Abwechslung für mich.
Abschließend möchte ich mich bei meiner Familie, meinen Freunden und ganz besonders bei Stefanie bedanken. Ihr habt immer an mich geglaubt, mich unterstützt und mir Rückhalt gegeben. Vielen Dank dafür.
87 88 LEBENSLAUF
Name: Andreas Stukenborg Geburtstag: 24. September 1979 Geburtsort: Vechta Familienstand: ledig Staatsangehörigkeit: deutsch
Schulbildung 1986 – 1990 Grundschule Goldenstedt 1990 – 1992 Orientierungsstufe Goldenstedt 1992 – 1999 Gymnasium Antonianum Vechta Abschluss: allgemeine Hochschulreife
Zivildienst 1999 – 2000 Zivildienst beim Naturschutz- und Informationszentrum Goldenstedt
Berufsausbildung 2000 – 2001 1. Ausbildungsjahr zum Landwirt auf dem Milchvieh- und Schweine- mastbetrieb Herbert Breithaupt, Ganderkesee 2001 – 2002 2. Ausbildungsjahr zum Landwirt auf dem Schweinezuchtbetrieb Heinz-Georg Feldhaus, Visbek 2002 – 2005 Studium der Agrarwissenschaften mit der Fachrichtung Tierproduktion an der Christian-Albrechts-Universität Kiel Abschluss: Bachelor of Science 2005 – 2007 Studium der Agrarwissenschaften mit der Fachrichtung Tierproduktion an der Christian-Albrechts-Universität Kiel Abschluss: Master of Science
Berufliche Tätigkeit 07/2002 – 10/2002 Mitarbeiter auf dem Schweinezuchtbetrieb Heinz-Georg Feldhaus, Visbek Seit Juni 2007 Wissenschaftlicher Mitarbeiter am Institut für Tierzucht und Tierhaltung der Christian-Albrechts-Universität zu Kiel bei Herrn Prof. Dr. J. Krieter
89